Machine Learning Street Talk
Mechanistic Interpretability - NEEL NANDA (DeepMind)
updated
This was Chollet's keynote talk at AGI-24, filmed in high-quality. We will be releasing a full interview with him shortly. A teaser clip from that is played in the intro!
Chollet introduces the Abstraction and Reasoning Corpus (ARC) as a benchmark for measuring AI progress towards human-like intelligence. He explains the concept of abstraction in AI systems and proposes combining deep learning with program synthesis to overcome current limitations. Chollet suggests that breakthroughs in AI might come from outside major tech labs and encourages researchers to explore new ideas in the pursuit of artificial general intelligence.
MLST is sponsored by Tufa Labs:
Are you interested in working on ARC and cutting-edge AI research with the MindsAI team (current ARC winners)?
Focus: ARC, LLMs, test-time-compute, active inference, system2 reasoning, and more.
Future plans: Expanding to complex environments like Warcraft 2 and Starcraft 2.
Interested? Apply for an ML research position: benjamin@tufa.ai
TOC
1. LLM Limitations and Intelligence Concepts
[00:00:00] 1.1 LLM Limitations and Composition
[00:12:05] 1.2 Intelligence as Process vs. Skill
[00:17:15] 1.3 Generalization as Key to AI Progress
2. ARC-AGI Benchmark and LLM Performance
[00:19:59] 2.1 Introduction to ARC-AGI Benchmark
[00:20:05] 2.2 Introduction to ARC-AGI and the ARC Prize
[00:23:35] 2.3 Performance of LLMs and Humans on ARC-AGI
3. Abstraction in AI Systems
[00:26:10] 3.1 The Kaleidoscope Hypothesis and Abstraction Spectrum
[00:30:05] 3.2 LLM Capabilities and Limitations in Abstraction
[00:32:10] 3.3 Value-Centric vs Program-Centric Abstraction
[00:33:25] 3.4 Types of Abstraction in AI Systems
4. Advancing AI: Combining Deep Learning and Program Synthesis
[00:34:05] 4.1 Limitations of Transformers and Need for Program Synthesis
[00:36:45] 4.2 Combining Deep Learning and Program Synthesis
[00:39:59] 4.3 Applying Combined Approaches to ARC Tasks
[00:44:20] 4.4 State-of-the-Art Solutions for ARC
Shownotes (new!): dropbox.com/scl/fi/i7nsyoahuei6np95lbjxw/CholletKeynote.pdf?rlkey=t3502kbov5exsdxhderq70b9i&st=1ca91ewz&dl=0
[0:01:15] Abstraction and Reasoning Corpus (ARC): AI benchmark (François Chollet)
arxiv.org/abs/1911.01547
[0:05:30] Monty Hall problem: Probability puzzle (Steve Selvin)
tandfonline.com/doi/abs/10.1080/00031305.1975.10479121
[0:06:20] LLM training dynamics analysis (Tirumala et al.)
arxiv.org/abs/2205.10770
[0:10:20] Transformer limitations on compositionality (Dziri et al.)
arxiv.org/abs/2305.18654
[0:10:25] Reversal Curse in LLMs (Berglund et al.)
arxiv.org/abs/2309.12288
[0:19:25] Measure of intelligence using algorithmic information theory (François Chollet)
arxiv.org/abs/1911.01547
[0:20:10] ARC-AGI: GitHub repository (François Chollet)
github.com/fchollet/ARC-AGI
[0:22:15] ARC Prize: $1,000,000+ competition (François Chollet)
arcprize.org
[0:33:30] System 1 and System 2 thinking (Daniel Kahneman)
amazon.com/Thinking-Fast-Slow-Daniel-Kahneman/dp/0374533555
[0:34:00] Core knowledge in infants (Elizabeth Spelke)
harvardlds.org/wp-content/uploads/2017/01/SpelkeKinzler07-1.pdf
[0:34:30] Embedding interpretive spaces in ML (Tennenholtz et al.)
arxiv.org/abs/2310.04475
[0:44:20] Hypothesis Search with LLMs for ARC (Wang et al.)
arxiv.org/abs/2309.05660
[0:44:50] Ryan Greenblatt's high score on ARC public leaderboard
arcprize.org
The conversation covers philosophical arguments about machine consciousness, including John Searle's famous Chinese Room thought experiment. They discuss different views on what's required for a system to truly understand or be conscious, touching on ideas from various philosophers and scientists.
The hosts also talk about the recent Nobel Prize awarded for work in deep learning, debating its merits and controversies. They touch on the recent Liron Doom Debates show which Duggar was just on, and at the end Ethics vs AI risk.
TOC:
1. Computational Foundations of AI
00:00:00 1.1 Turing Completeness and Computational Limits of Language Models
00:06:37 1.2 Finite State Automata vs. Turing Machines in AI
00:13:16 1.3 Challenges in Training Turing-Complete Systems
00:17:23 1.4 Mapping Turing Machine Programs to Language Models
00:20:41 1.5 Future Directions: Hybrid Systems and Novel Programming Approaches
2. AI Consciousness and Understanding
00:31:30 2.1 Chinese Room Argument and AI Consciousness
00:41:15 2.2 Searle's Views on Physical Realization and Consciousness
00:50:30 2.3 Emergence and Computational Limitations in Cognitive Science
00:59:55 2.4 Friston's Theory on Self-Awareness in Machines
01:04:25 2.5 Causal Structures and Understanding in AI Systems
01:06:50 2.6 Concept Role Semantics and Language Models
01:09:41 2.7 Consciousness and Computational Theories
3. AI Impact and Ethics
01:21:38 3.1 Deep Learning and the Nobel Prize for Hinton
01:29:56 3.2 AI Harm and Existential Risk
01:34:05 3.3 Balancing AI Ethics and Practical Policy
Refs:
Keith/Liron discussion on Doom Debates (must watch!)
youtube.com/watch?v=4v-Qh3JQ4Jc
Liron's show was in response to our previous "steak house" chat-
Is o1 reasoning?
youtube.com/watch?v=nO6sDk6vO0g&t=0s
Minds brains and programs (Searle)
https://home.csulb.edu/~cwallis/382/readings/482/searle.minds.brains.programs.bbs.1980.pdf
Searle Google talk
youtube.com/watch?v=rHKwIYsPXLg
J. Mark Bishop on MLST
youtube.com/watch?v=_KVAzAzO5HU
Chinese room argument
https://plato.stanford.edu/entries/chinese-room/
Dancing with pixies - J. Mark Bishop
philarchive.org/rec/BISDWP-2
Artificial Intelligence Is Stupid and Causal Reasoning Will Not Fix It - Bishop
frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2020.513474/full
Deconstructing the AI Myth: Fallacies and Harms of Algorithmification (Dagmar Monnett)
researchgate.net/publication/382802495_Deconstructing_the_AI_Myth_Fallacies_and_Harms_of_Algorithmification
Nestedly Recursive Functions (Stephen Wolfram)
writings.stephenwolfram.com/2024/09/nestedly-recursive-functions
Measure of intelligence - Chollet
arxiv.org/abs/1911.01547
What is the philosophy of information - Floridi
researchgate.net/publication/225070243_What_Is_the_Philosophy_of_Information
Godel, Escher, Bach: An Eternal Golden Braid
amazon.co
.uk/Godel-Escher-Bach-Eternal-Golden/dp/0465026567
Shownotes/transcript/refs (autogenerated)
dropbox.com/scl/fi/jc6qwn7m2i69t13bo2hlm/TimKeithPSH2.pdf?rlkey=wtegece4e32v4xhvp9expnpcq&st=1tm6868m&dl=0
x
Recorded Friday 11th Oct 2024
Zhang highlights their implementation of Retrieval-Augmented Generation in healthcare, significantly reducing doctor preparation time. He explores the shift from monolithic AI models to heterogeneous systems and the importance of improving various AI system components. Zhang shares insights on using synthetic data to teach models reasoning, the democratization of software development through AI, and how his gaming skills transfer to running an AI company.
He advises young developers to fully embrace AI technologies and offers perspectives on AI reliability, potential risks, and future model architectures.
cohere.com
ivanzhang.ca
https://x.com/1vnzh
TOC:
00:00:00 Intro
00:03:20 AI & Language Model Evolution
00:06:09 Future AI Apps & Development
00:09:29 Impact on Software Dev Practices
00:13:03 Philosophical & Societal Implications
00:16:30 Compute Efficiency & RAG
00:20:39 Adoption Challenges & Solutions
00:22:30 GPU Optimization & Kubernetes Limits
00:24:16 Cohere's Implementation Approach
00:28:13 Gaming's Professional Influence
00:34:45 Transformer Optimizations
00:36:45 Future Models & System-Level Focus
00:39:20 Inference-Time Computation & Reasoning
00:42:05 Capturing Human Thought in AI
00:43:15 Research, Hiring & Developer Advice
REFS:
00:02:40 The Transformer architecture, arxiv.org/abs/1706.03762
00:03:22 The Innovator's Dilemma, amazon.com/Innovators-Dilemma-Technologies-Management-Innovation/dp/1633691780
00:09:15 The actor model, en.wikipedia.org/wiki/Actor_model
00:14:35 John Searle's Chinese Room Argument, https://plato.stanford.edu/entries/chinese-room/
00:18:00 Retrieval-Augmented Generation, arxiv.org/abs/2005.11401
00:18:40 Retrieval-Augmented Generation, docs.cohere.com/v2/docs/retrieval-augmented-generation-rag
00:35:39 Let’s Verify Step by Step, arxiv.org/pdf/2305.20050
00:39:20 Adaptive Inference-Time Compute, arxiv.org/abs/2410.02725
00:43:20 Ryan Greenblatt ARC entry, redwoodresearch.substack.com/p/getting-50-sota-on-arc-agi-with-gpt
Disclaimer: This show is part of our Cohere partnership series.
TOC:
00:00:00 Introduction to Open-Ended AI and Key Concepts
00:01:37 Tim Rocktäschel's Background and Research Focus
00:06:25 Defining Open-Endedness in AI Systems
00:10:39 Subjective Nature of Interestingness and Learnability
00:16:22 Open-Endedness in Practice: Examples and Limitations
00:17:50 Assessing Novelty in Open-ended AI Systems
00:20:05 Adversarial Attacks and AI Robustness
00:24:05 Rainbow Teaming and LLM Safety
00:25:48 Open-ended Research Approaches in AI
00:29:05 Balancing Long-term Vision and Exploration in AI Research
00:37:25 LLMs in Program Synthesis and Open-Ended Learning
00:37:55 Transition from Human-Based to Novel AI Strategies
00:39:00 Expanding Context Windows and Prompt Evolution
00:40:17 AI Intelligibility and Human-AI Interfaces
00:46:04 Self-Improvement and Evolution in AI Systems
Show notes (New!) dropbox.com/scl/fi/5avpsyz8jbn4j1az7kevs/TimR.pdf?rlkey=pqjlcqbtm3undp4udtgfmie8n&st=x50u1d1m&dl=0
REFS:
00:01:47 - UCL DARK Lab (Rocktäschel) - AI research lab focusing on RL and open-ended learning - ucldark.com
00:02:31 - GENIE (Bruce) - Generative interactive environment from unlabelled videos - arxiv.org/abs/2402.15391
00:02:42 - Promptbreeder (Fernando) - Self-referential LLM prompt evolution - arxiv.org/abs/2309.16797
00:03:05 - Picbreeder (Secretan) - Collaborative online image evolution - dl.acm.org/doi/10.1145/1357054.1357328
00:03:14 - Why Greatness Cannot Be Planned (Stanley) - Book on open-ended exploration - amazon.com/Why-Greatness-Cannot-Planned-Objective/dp/3319155237
00:04:36 - NetHack Learning Environment (Küttler) - RL research in procedurally generated game - arxiv.org/abs/2006.13760
00:07:35 - Open-ended learning (Clune) - AI systems for continual learning and adaptation - arxiv.org/abs/1905.10985
00:07:35 - OMNI (Zhang) - LLMs modeling human interestingness for exploration - arxiv.org/abs/2306.01711
00:10:42 - Observer theory (Wolfram) - Computationally bounded observers in complex systems - writings.stephenwolfram.com/2023/12/observer-theory
00:15:25 - Human-Timescale Adaptation (Rocktäschel) - RL agent adapting to novel 3D tasks - arxiv.org/abs/2301.07608
00:16:15 - Open-Endedness for AGI (Hughes) - Importance of open-ended learning for AGI - arxiv.org/abs/2406.04268
00:16:35 - POET algorithm (Wang) - Open-ended approach to generate and solve challenges - arxiv.org/abs/1901.01753
00:17:20 - AlphaGo (Silver) - AI mastering the game of Go - https://deepmind.google/technologies/alphago/
00:20:35 - Adversarial Go attacks (Dennis) - Exploiting weaknesses in Go AI systems - ifaamas.org/Proceedings/aamas2024/pdfs/p1630.pdf
00:22:00 - Levels of AGI (Morris) - Framework for categorizing AGI progress - arxiv.org/abs/2311.02462
00:24:30 - Rainbow Teaming (Samvelyan) - LLM-based adversarial prompt generation - arxiv.org/abs/2402.16822
00:25:50 - Why Greatness Cannot Be Planned (Stanley) - 'False compass' and 'stepping stone collection' concepts - amazon.com/Why-Greatness-Cannot-Planned-Objective/dp/3319155237
00:27:45 - AI Debate (Khan) - Improving LLM truthfulness through debate - https://proceedings.mlr.press/v235/khan24a.html
00:29:40 - Gemini (Google DeepMind) - Advanced multimodal AI model - https://deepmind.google/technologies/gemini/
00:30:15 - How to Take Smart Notes (Ahrens) - Effective note-taking methodology - amazon.com/How-Take-Smart-Notes-Nonfiction/dp/1542866502
00:35:05 - Voyager (Wang) - Open-ended embodied agent using GPT-4 in Minecraft - arxiv.org/abs/2305.16291
00:38:00 - AlphaGo Nature paper (Silver) - Deep neural networks and tree search for Go - nature.com/articles/nature16961
00:38:05 - AlphaStar (Vinyals) - AI achieving grandmaster level in StarCraft II - nature.com/articles/s41586-019-1724-z
00:42:00 - The Beginning of Infinity (Deutsch) - Book on explanations and scientific progress - amazon.com/Beginning-Infinity-Explanations-Transform-World/dp/0143121359
00:43:30 - AI model collapse (Shumailov) - Risks of training on AI-generated content - nature.com/articles/s41586-024-07566-y
00:48:35 - Chain-of-Thought Prompting (Wei) - Improving LLM reasoning through prompting - arxiv.org/abs/2201.11903
00:49:35 - Self-improving neural networks (Schmidhuber) - Early work on self-referential networks - http://people.idsia.ch/~juergen/selfrefnn.html
00:54:45 - UCL DARK Lab (UCL Computer Science) - RL and Deep Learning research group - ucl.ac.uk/computer-science/news-and-events/inaugural-lecture-series/inaugural-lecture-professor-tim-rocktaschel
TOC:
[00:00:00] AGI Timeline Predictions and Development Speed
[00:00:45] Limitations of Language Models in AGI Development
[00:02:18] Current State and Trends in AI Research and Development
[00:09:02] Emergent Reasoning Capabilities and Limitations of LLMs
[00:18:15] Neuro-Symbolic Approaches and the Future of AI Systems
[00:20:00] Evolutionary Algorithms and LLMs in Creative Tasks
[00:21:25] Symbolic vs. Sub-Symbolic Approaches in AI
[00:28:05] Language as Internal Thought and External Communication
[00:30:20] AGI Development and Goal-Directed Behavior
[00:35:51] Consciousness and AI: Expanding States of Experience
[00:48:50] AI Regulation: Challenges and Approaches
[00:55:35] Challenges in AI Regulation
[00:59:20] AI Alignment and Ethical Considerations
[01:09:15] AGI Development Timeline Predictions
[01:12:40] OpenCog Hyperon and AGI Progress
[01:17:48] Transhumanism and Resource Allocation Debate
[01:20:12] Cultural Perspectives on Transhumanism
[01:23:54] AGI and Post-Scarcity Society
[01:31:35] Challenges and Implications of AGI Development
New! PDF Show notes: dropbox.com/scl/fi/fyetzwgoaf70gpovyfc4x/BenGoertzel.pdf?rlkey=pze5dt9vgf01tf2wip32p5hk5&st=svbcofm3&dl=0
Refs:
00:00:15 Ray Kurzweil's AGI timeline prediction, Ray Kurzweil, en.wikipedia.org/wiki/Technological_singularity
00:01:45 Ben Goertzel: SingularityNET founder, Ben Goertzel, singularitynet.io
00:02:35 AGI Conference series, AGI Conference Organizers, agi-conf.org/2024
00:03:55 Ben Goertzel's contributions to AGI, Wikipedia contributors, en.wikipedia.org/wiki/Ben_Goertzel
00:11:05 Chain-of-Thought prompting, Subbarao Kambhampati, arxiv.org/abs/2405.04776
00:11:35 Algorithmic information content, Pieter Adriaans, https://plato.stanford.edu/entries/information-entropy/
00:12:10 Turing completeness in neural networks, Various contributors, https://plato.stanford.edu/entries/turing-machine/
00:16:15 AlphaGeometry: AI for geometry problems, Trieu, Li, et al., nature.com/articles/s41586-023-06747-5
00:18:25 Shane Legg and Ben Goertzel's collaboration, Shane Legg, en.wikipedia.org/wiki/Shane_Legg
00:20:00 Evolutionary algorithms in music generation, Yanxu Chen, arxiv.org/html/2409.03715v1
00:22:00 Peirce's theory of semiotics, Charles Sanders Peirce, https://plato.stanford.edu/entries/peirce-semiotics/
00:28:10 Chomsky's view on language, Noam Chomsky, chomsky.info/1983____
00:34:05 Greg Egan's 'Diaspora', Greg Egan, amazon.co.uk/Diaspora-post-apocalyptic-thriller-perfect-MIRROR/dp/0575082097
00:40:35 'The Consciousness Explosion', Ben Goertzel & Gabriel Axel Montes, amazon.com/Consciousness-Explosion-Technological-Experiential-Singularity/dp/B0D8C7QYZD
00:41:55 Ray Kurzweil's books on singularity, Ray Kurzweil, amazon.com/Singularity-Near-Humans-Transcend-Biology/dp/0143037889
00:50:50 California AI regulation bills, California State Senate, sd18.senate.ca.gov/news/senate-unanimously-approves-senator-padillas-artificial-intelligence-package
00:56:40 Limitations of Compute Thresholds, Sara Hooker, arxiv.org/abs/2407.05694
00:56:55 'Taming Silicon Valley', Gary F. Marcus, penguinrandomhouse.com/books/768076/taming-silicon-valley-by-gary-f-marcus
01:09:15 Kurzweil's AGI prediction update, Ray Kurzweil, theguardian.com/technology/article/2024/jun/29/ray-kurzweil-google-ai-the-singularity-is-nearer
01:14:45 OpenCog Hyperon framework, Ben Goertzel et al., arxiv.org/abs/2310.18318
01:18:25 Malnutrition in Ethiopia, Abriham Shiferaw Areba, frontiersin.org/journals/nutrition/articles/10.3389/fnut.2024.1403591/full
01:18:40 Transhumanism ethical debate, Nick Bostrom, nickbostrom.com/papers/history.pdf
Key points discussed:
- Using debate between language models to improve truthfulness in AI responses
- Scalable oversight for supervising AI models beyond human-level intelligence
- The relationship between intelligence and agency in AI systems
- Challenges in AI safety and alignment
- The potential for a Cambrian explosion in human-like intelligent systems
The discussion also explored broader topics:
- The wisdom of crowds vs. expert knowledge in machine learning debates
- Deceptive alignment and reward tampering in AI systems
- Open-ended AI systems and their implications for development and safety
- The space of possible minds and defining superintelligence
- Cultural evolution and memetics in understanding intelligence
Akbir Khan:
https://x.com/akbirkhan
https://akbir.dev/
Show notes and transcript: dropbox.com/scl/fi/sjekivbg3ok6qugsv2p1u/AkbirKhan.pdf?rlkey=ewiyvq0aq7mjvql4u7os0jos2&st=vblhp7af&dl=0
TOC (*) are best bits
00:00:00 1. Intro: AI alignment and debate techniques for truthful responses *
00:05:00 2. Scalable oversight and hidden information settings
00:10:05 3. AI agency, intelligence, and progress *
00:15:00 4. Base models, RL training, and instrumental goals
00:25:11 5. Deceptive alignment and RL challenges in AI *
00:30:12 6. Open-ended AI systems and future directions
00:35:34 7. Deception, superintelligence, and the space of possible minds *
00:40:00 8. Cultural evolution, memetics, and intelligence measurement
References:
1. [00:00:40] Akbir Khan et al. ICML 2024 Best Paper: "Debating with More Persuasive LLMs Leads to More Truthful Answers"
arxiv.org/html/2402.06782v3
2. [00:03:28] Yann LeCun on machine learning debates
youtube.com/watch?v=OKkEdTchsiE
3. [00:06:05] OpenAI's Superalignment team
openai.com/index/introducing-superalignment
4. [00:08:10] Sam Bowman on scalable oversight in AI systems
arxiv.org/abs/2211.03540
5. [00:10:35] Sam Bowman on the sandwich protocol
alignmentforum.org/posts/nekLYqbCEBDEfbLzF/artificial-sandwiching-when-can-we-test-scalable-alignment
6. [00:14:35] Janus' article on "Simulators" and LLMs
lesswrong.com/posts/vJFdjigzmcXMhNTsx/simulators
7. [00:16:35] Thomas Suddendorf's book "The Gap: The Science of What Separates Us from Other Animals"
https://www.amazon.in/GAP-Science-Separates-Other-Animals/dp/0465030149
8. [00:19:10] DeepMind on responsible AI
https://deepmind.google/about/responsibility-safety/
9. [00:20:50] Technological singularity
en.wikipedia.org/wiki/Technological_singularity
10. [00:21:30] Eliezer Yudkowsky on FOOM (Fast takeoff)
intelligence.org/files/AIFoomDebate.pdf
11. [00:21:45] Sammy Martin on recursive self-improvement in AI
alignmentforum.org/posts/5WECpYABCT62TJrhY/will-ai-undergo-discontinuous-progress
12. [00:24:25] LessWrong community
lesswrong.com
13. [00:24:35] Nora Belrose on AI alignment and deception
lesswrong.com/posts/YsFZF3K9tuzbfrLxo/counting-arguments-provide-no-evidence-for-ai-doom
14. [00:25:35] Evan Hubinger on deceptive alignment in AI systems
lesswrong.com/posts/zthDPAjh9w6Ytbeks/deceptive-alignment
15. [00:26:50] Anthropic's article on reward tampering in language models
anthropic.com/research/reward-tampering
16. [00:32:35] Kenneth Stanley's work on open-endedness in AI
amazon.co.uk/Why-Greatness-Cannot-Planned-Objective/dp/3319155237
17. [00:34:58] Ryan Greenblatt, Buck Shlegeris et al. on AI safety protocols
arxiv.org/pdf/2312.06942
18. [00:37:20] Aaron Sloman's concept of 'the space of possible minds'
cs.bham.ac.uk/research/projects/cogaff/sloman-space-of-minds-84.pdf
19. [00:38:25] François Chollet on defining and measuring intelligence in AI
arxiv.org/abs/1911.01547
20. [00:42:30] Richard Dawkins on memetics
amazon.co.uk/Selfish-Gene-Richard-Dawkins/dp/0192860925
21. [00:42:45] Jonathan Cook et al. on Artificial Generational Intelligence
arxiv.org/abs/2406.00392
22. [00:45:00] Peng on determinants of cryptocurrency pricing
emerald.com/insight/content/doi/10.1108/CAFR-05-2023-0053/full/html
Marcus is worried about tech companies putting profits before people. He thinks AI could make fake news and privacy issues even worse. He's also concerned that a few big tech companies have too much power. Looking ahead, Marcus believes the AI hype will die down as reality sets in. He wants to see AI developed in smarter, more responsible ways. His message to the public? We need to speak up and demand better AI before it's too late.
Buy Taming Silicon Valley:
amzn.to/3XTlC5s
Gary Marcus:
garymarcus.substack.com
https://x.com/GaryMarcus
Interviewer:
Dr. Tim Scarfe
(Refs in top comment)
TOC
[00:00:00] AI Flaws, Improvements & Industry Critique
[00:16:29] AI Safety Theater & Image Generation Issues
[00:23:49] AI's Lack of World Models & Human-like Understanding
[00:31:09] LLMs: Superficial Intelligence vs. True Reasoning
[00:34:45] AI in Specialized Domains: Chess, Coding & Limitations
[00:42:10] AI-Generated Code: Capabilities & Human-AI Interaction
[00:48:10] AI Regulation: Industry Resistance & Oversight Challenges
[00:54:55] Copyright Issues in AI & Tech Business Models
[00:57:26] AI's Societal Impact: Risks, Misinformation & Ethics
[01:23:14] AI X-risk, Alignment & Moral Principles Implementation
[01:37:10] Persistent AI Flaws: System Limitations & Architecture Challenges
[01:44:33] AI Future: Surveillance Concerns, Economic Challenges & Neuro-Symbolic AI
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Key points discussed:
The limitations of cortex-centric approaches to consciousness studies.
Evidence from decorticated animals and hydranencephalic children supporting the brainstem's role in consciousness.
The relationship between homeostasis, the free energy principle, and consciousness.
Critiques of behaviorism and modern theories of consciousness.
The importance of subjective experience in understanding brain function.
The discussion also explored broader topics:
The potential impact of affect-based theories on AI development.
The role of the SEEKING system in exploration and learning.
Connections between neuroscience, psychoanalysis, and philosophy of mind.
Challenges in studying consciousness and the limitations of current theories.
Mark Solms:
https://neuroscience.uct.ac.za/contacts/mark-solms
Show notes and transcript: dropbox.com/scl/fo/roipwmnlfmwk2e7kivzms/ACjZF-VIGC2-Suo30KcwVV0?rlkey=53y8v2cajfcgrf17p1h7v3suz&st=z8vu81hn&dl=0
TOC (*) are best bits
00:00:00 1. Intro: Challenging vision-centric approaches to consciousness *
00:02:20 2. Evidence from decorticated animals and hydranencephalic children *
00:07:40 3. Emotional responses in hydranencephalic children
00:10:40 4. Brainstem stimulation and affective states
00:15:00 5. Brainstem's role in generating affective consciousness *
00:21:50 6. Dual-aspect monism and the mind-brain relationship
00:29:37 7. Information, affect, and the hard problem of consciousness *
00:37:25 8. Wheeler's participatory universe and Chalmers' theories
00:48:51 9. Homeostasis, free energy principle, and consciousness *
00:59:25 10. Affect, voluntary behavior, and decision-making
01:05:45 11. Psychoactive substances, REM sleep, and consciousness research
01:12:14 12. Critiquing behaviorism and modern consciousness theories *
01:24:25 13. The SEEKING system and exploration in neuroscience
Refs:
1. Mark Solms' book "The Hidden Spring" [00:20:34] (MUST READ!)
amzn.to/3XyETb3
2. Karl Friston's free energy principle [00:03:50]
nature.com/articles/nrn2787
3. Hydranencephaly condition [00:07:10]
en.wikipedia.org/wiki/Hydranencephaly
4. Periaqueductal gray (PAG) [00:08:57]
en.wikipedia.org/wiki/Periaqueductal_gray
5. Positron Emission Tomography (PET) [00:13:52]
en.wikipedia.org/wiki/Positron_emission_tomography
6. Paul MacLean's triune brain theory [00:03:30]
en.wikipedia.org/wiki/Triune_brain
7. Baruch Spinoza's philosophy of mind [00:23:48]
https://plato.stanford.edu/entries/spinoza-epistemology-mind
8. Claude Shannon's "A Mathematical Theory of Communication" [00:32:15]
https://people.math.harvard.edu/~ctm/home/text/others/shannon/entropy/entropy.pdf
9. Francis Crick's "The Astonishing Hypothesis" [00:39:57]
en.wikipedia.org/wiki/The_Astonishing_Hypothesis
10. Frank Jackson's Knowledge Argument [00:40:54]
https://plato.stanford.edu/entries/qualia-knowledge/
11. Mesolimbic dopamine system [01:11:51]
en.wikipedia.org/wiki/Mesolimbic_pathway
12. Jaak Panksepp's SEEKING system [01:25:23]
en.wikipedia.org/wiki/Jaak_Panksepp#Affective_neuroscience
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmented generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Key topics covered:
- Origins and evolution of Retrieval Augmented Generation (RAG)
- Challenges in evaluating RAG systems and language models
- Human-AI collaboration in research and knowledge work
- Word embeddings and the progression to modern language models
- Dense vs sparse retrieval methods in information retrieval
The discussion also explored broader implications and applications:
- Balancing faithfulness and fluency in RAG systems
- User interface design for AI-augmented research tools
- The journey from chemistry to AI research
- Challenges in enterprise search compared to web search
- The importance of data quality in training AI models
Patrick Lewis: patricklewis.io
Cohere Command Models, check them out - they are amazing for RAG!
cohere.com/command
TOC
00:00:00 1. Intro to RAG
00:05:30 2. RAG Evaluation: Poll framework & model performance
00:12:55 3. Data Quality: Cleanliness vs scale in AI training
00:15:13 4. Human-AI Collaboration: Research agents & UI design
00:22:57 5. RAG Origins: Open-domain QA to generative models
00:30:18 6. RAG Challenges: Info retrieval, tool use, faithfulness
00:42:01 7. Dense vs Sparse Retrieval: Techniques & trade-offs
00:47:02 8. RAG Applications: Grounding, attribution, hallucination prevention
00:54:04 9. UI for RAG: Human-computer interaction & model optimization
00:59:01 10. Word Embeddings: Word2Vec, GloVe, and semantic spaces
01:06:43 11. Language Model Evolution: BERT, GPT, and beyond
01:11:38 12. AI & Human Cognition: Sequential processing & chain-of-thought
Refs:
1. Retrieval Augmented Generation (RAG) paper / Patrick Lewis et al. [00:27:45]
arxiv.org/abs/2005.11401
2. LAMA (LAnguage Model Analysis) probe / Petroni et al. [00:26:35]
arxiv.org/abs/1909.01066
3. KILT (Knowledge Intensive Language Tasks) benchmark / Petroni et al. [00:27:05]
arxiv.org/abs/2009.02252
4. Word2Vec algorithm / Tomas Mikolov et al. [01:00:25]
arxiv.org/abs/1301.3781
5. GloVe (Global Vectors for Word Representation) / Pennington et al. [01:04:35]
https://nlp.stanford.edu/projects/glove/
6. BERT (Bidirectional Encoder Representations from Transformers) / Devlin et al. [01:08:00]
arxiv.org/abs/1810.04805
7. 'The Language Game' book / Nick Chater and Morten H. Christiansen [01:11:40]
amzn.to/4grEUpG
Disclaimer: This is the sixth video from our Cohere partnership. We were not told what to say in the interview. Filmed in London in June 2024.
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
We edited about an hour off this conversation, see full one on patreon - patreon.com/posts/tim-and-keith-on-112091132
TOC:
00:00:00 1. Introduction and AI hype cycles
00:02:09 2. Computational limits of AI systems
00:03:57 3. Neural Networks vs. Turing Machines
00:11:55 4. Computational models in AI
00:13:03 5. What is Reasoning?
00:21:08 6. Chain-of-thought prompting
00:26:02 7. AI code generation and complexity
00:34:24 8. AI assistance vs. human problem-solving
00:35:04 9. Limitations of AI in reasoning and problem-solving
00:46:27 10. Knowledge acquisition and inference in AI
00:53:36 11. Comparing AI and human reasoning capabilities
00:58:58 12. LLMs as cognitive tools
01:00:32 13. Testing o1-preview on a logic puzzle
01:20:48 14. AI-assisted coding: strengths and limitations
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Genie's approach to learning interactive environments, balancing compression and fidelity.
The use of latent action models and VQE models for video processing and tokenization.
Challenges in maintaining action consistency across frames and integrating text-to-image models.
Evaluation metrics for AI-generated content, such as FID and PS&R diff metrics.
The discussion also explored broader implications and applications:
The potential impact of AI video generation on content creation jobs.
Applications of Genie in game generation and robotics.
The use of foundation models in robotics and the differences between internet video data and specialized robotics data.
Challenges in mapping AI-generated actions to real-world robotic actions.
Ashley Edwards: ashedwards.github.io
TOC (*) are best bits
00:00:00 1. Intro to Genie & Brave Search API: Trade-offs & limitations *
00:02:26 2. Genie's Architecture: Latent action, VQE, video processing *
00:05:06 3. Genie's Constraints: Frame consistency & image model integration
00:07:26 4. Evaluation: FID, PS&R diff metrics & latent induction methods
00:09:44 5. AI Video Gen: Content creation impact, depth & parallax effects
00:11:39 6. Model Scaling: Training data impact & computational trade-offs
00:13:50 7. Game & Robotics Apps: Gamification & action mapping challenges *
00:16:16 8. Robotics Foundation Models: Action space & data considerations *
00:19:18 9. Mask-GPT & Video Frames: Real-time optimization, RL from videos
00:20:34 10. Research Challenges: AI value, efficiency vs. quality, safety
00:24:20 11. Future Dev: Efficiency improvements & fine-tuning strategies
Refs:
1. Genie (learning interactive environments from videos) / Ashley and DM collegues [00:01]
arxiv.org/abs/2402.15391
2. VQ-VAE (Vector Quantized Variational Autoencoder) / Aaron van den Oord, Oriol Vinyals, Koray Kavukcuoglu [02:43]
arxiv.org/abs/1711.00937
3. FID (Fréchet Inception Distance) metric / Martin Heusel et al. [07:37]
arxiv.org/abs/1706.08500
4. PS&R (Precision and Recall) metric / Mehdi S. M. Sajjadi et al. [08:02]
arxiv.org/abs/1806.00035
5. Vision Transformer (ViT) architecture / Alexey Dosovitskiy et al. [12:14]
arxiv.org/abs/2010.11929
6. Genie (robotics foundation models) / Google DeepMind [17:34]
https://deepmind.google/research/publications/60474/
7. Chelsea Finn's lab work on robotics datasets / Chelsea Finn [17:38]
https://ai.stanford.edu/~cbfinn/
8. Imitation from observation in reinforcement learning / YuXuan Liu [20:58]
arxiv.org/abs/1707.03374
9. Waymo's autonomous driving technology / Waymo [22:38]
waymo.com
10. Gen3 model release by Runway / Runway [23:48]
runwayml.com
11. Classifier-free guidance technique / Jonathan Ho and Tim Salimans [24:43]
arxiv.org/abs/2207.12598
* Cohere focuses on pragmatic, efficient models tailored for business applications rather than pursuing the largest possible models.
* They offer flexible deployment options, from cloud services to on-premises installations, to meet diverse enterprise needs.
* Retrieval-augmented generation (RAG) is highlighted as a critical capability, allowing models to leverage enterprise data securely.
* Cohere emphasizes model customization, fine-tuning, and tools like reranking to optimize performance for specific use cases.
* The company has seen significant growth, transitioning from developer-focused to enterprise-oriented services.
* Major customers like Oracle, Fujitsu, and TD Bank are using Cohere's models across various applications, from HR to finance.
* Baji predicts a surge in enterprise AI adoption over the next 12-18 months as more companies move from experimentation to production.
* He emphasizes the importance of trust, security, and verifiability in enterprise AI applications.
The interview provides insights into Cohere's strategy, technology, and vision for the future of enterprise AI adoption.
linkedin.com/in/saurabhbaji
https://x.com/sbaji
cohere.com
cohere.com/business
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
TOC (*) are best bits
00:00:00 1. Introduction and Background
00:04:24 2. Cloud Infrastructure and LLM Optimization
00:06:43 2.1 Model deployment and fine-tuning strategies *
00:09:37 3. Enterprise AI Deployment Strategies
00:11:10 3.1 Retrieval-augmented generation in enterprise environments *
00:13:40 3.2 Standardization vs. customization in cloud services *
00:18:20 4. AI Model Evaluation and Deployment
00:18:20 4.1 Comprehensive evaluation frameworks *
00:21:20 4.2 Key components of AI model stacks *
00:25:50 5. Retrieval Augmented Generation (RAG) in Enterprise
00:32:10 5.1 Pragmatic approach to RAG implementation *
00:33:45 6. AI Agents and Tool Integration
00:33:45 6.1 Leveraging tools for AI insights *
00:35:30 6.2 Agent-based AI systems and diagnostics *
00:42:55 7. AI Transparency and Reasoning Capabilities
00:49:10 8. AI Model Training and Customization
00:57:10 9. Enterprise AI Model Management
01:02:10 9.1 Managing AI model versions for enterprise customers *
01:04:30 9.2 Future of language model programming *
01:06:10 10. AI-Driven Software Development
01:06:10 10.1 AI bridging human expression and task achievement *
01:08:00 10.2 AI-driven virtual app fabrics in enterprise *
01:13:33 11. Future of AI and Enterprise Applications
01:21:55 12. Cohere's Customers and Use Cases
01:21:55 12.1 Cohere's growth and enterprise partnerships *
01:27:14 12.2 Diverse customers using generative AI *
01:27:50 12.3 Industry adaptation to generative AI *
01:29:00 13. Technical Advantages of Cohere Models
01:29:00 13.1 Handling large context windows *
01:29:40 13.2 Low latency impact on developer productivity *
Disclaimer: This is the fifth video from our Cohere partnership. We were not told what to say in the interview, and didn't edit anything out from the interview. Filmed in Seattle in Aug 2024.
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
The interview with David Hanson covers:
The importance of incorporating biological drives and compassion into AI systems
Hanson's concept of "existential pattern ethics" as a basis for AI morality
The potential for AI to enhance human intelligence and wisdom
Challenges in developing artificial general intelligence (AGI)
The need to democratize AI technologies globally
Potential future advancements in human-AI integration and their societal impacts
Concerns about technological augmentation exacerbating inequality
The role of ethics in guiding AI development and deployment
Hanson advocates for creating AI systems that embody the best aspects of humanity while surpassing current human limitations, aiming for "super wisdom" rather than just artificial super intelligence.
David Hanson:
hansonrobotics.com/david-hanson
youtube.com/watch?v=9u1O954cMmE
TOC
1. Introduction and Background [00:00:00]
1.1. David Hanson's interdisciplinary background [0:01:49]
1.2. Introduction to Sofia, the realistic robot [0:03:27]
2. Human Cognition and AI [0:03:50]
2.1. Importance of social interaction in cognition [0:03:50]
2.2. Compassion as distinguishing factor [0:05:55]
2.3. AI augmenting human intelligence [0:09:54]
3. Developing Human-like AI [0:13:17]
3.1. Incorporating biological drives in AI [0:13:17]
3.2. Creating AI with agency [0:20:34]
3.3. Implementing flexible desires in AI [0:23:23]
4. Ethics and Morality in AI [0:27:53]
4.1. Enhancing humanity through AI [0:27:53]
4.2. Existential pattern ethics [0:30:14]
4.3. Expanding morality beyond restrictions [0:35:35]
5. Societal Impact of AI [0:38:07]
5.1. AI adoption and integration [0:38:07]
5.2. Democratizing AI technologies [0:38:32]
5.3. Human-AI integration and identity [0:43:37]
6. Future Considerations [0:50:03]
6.1. Technological augmentation and inequality [0:50:03]
6.2. Emerging technologies for mental health [0:50:32]
6.3. Corporate ethics in AI development [0:52:26]
This was filmed at AGI-24
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
We discuss abstract concepts like collective intelligence, the importance of embodiment in understanding the world, and how we acquire and process knowledge. Spivak shares his thoughts on creativity, discussing where it comes from and how it might be modeled mathematically.
A significant portion of the discussion focuses on the impact of artificial intelligence on human thinking and its potential role in the evolution of intelligence. Spivak also touches on the importance of language, particularly written language, in transmitting knowledge and shaping our understanding of the world.
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/The-Fabric-of-Knowledge---David-Spivak-e2o220h
David Spivak:
http://www.dspivak.net
TOC:
00:00:00 Introduction to category theory and functors
00:04:40 Collective intelligence and sense-making
00:09:54 Embodiment and physical concepts in knowledge acquisition
00:16:23 Creativity, open-endedness, and AI's impact on thinking
00:25:46 Modeling creativity and the evolution of intelligence
00:36:04 Evolution, optimization, and the significance of AI
00:44:14 Written language and its impact on knowledge transmission
REFS:
Mike Levin's work
scholar.google.com/citations?user=luouyakAAAAJ&hl=en
Eric Smith's videos on complexity and early life
youtube.com/watch?v=SpJZw-68QyE
Richard Dawkins' book "The Selfish Gene"
amzn.to/3X73X8w
Carl Sagan's statement about the cosmos knowing itself
amzn.to/3XhPruK
Herbert Simon's concept of "satisficing"
https://plato.stanford.edu/entries/bounded-rationality/
DeepMind paper on open-ended systems
arxiv.org/abs/2406.04268
Karl Friston's work on active inference
https://direct.mit.edu/books/oa-monograph/5299/Active-InferenceThe-Free-Energy-Principle-in-Mind
MIT category theory lectures by David Spivak (available on the Topos Institute channel)
youtube.com/watch?v=UusLtx9fIjs
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
TOC
00:00:00 Intro
00:03:38 Reasoning
00:13:09 Potential AI Breakthroughs Reducing Computation Needs
00:20:39 Memorization vs. Generalization in AI
00:25:19 Approach to the ARC Challenge
00:29:10 Perceptions of Chat GPT and AGI
00:58:45 Abstract Principles of Jurgen's Approach
01:04:17 Analogical Reasoning and Compression
01:05:48 Breakthroughs in 1991: the P, the G, and the T in ChatGPT and Generative AI
01:15:50 Use of LSTM in Language Models by Tech Giants
01:21:08 Neural Network Aspect Ratio Theory
01:26:53 Reinforcement Learning Without Explicit Teachers
Refs:
★ "Annotated History of Modern AI and Deep Learning" (2022 survey by Schmidhuber):
★ Chain Rule For Backward Credit Assignment (Leibniz, 1676)
★ First Neural Net / Linear Regression / Shallow Learning (Gauss & Legendre, circa 1800)
★ First 20th Century Pioneer of Practical AI (Quevedo, 1914)
★ First Recurrent NN (RNN) Architecture (Lenz, Ising, 1920-1925)
★ AI Theory: Fundamental Limitations of Computation and Computation-Based AI (Gödel, 1931-34)
★ Unpublished ideas about evolving RNNs (Turing, 1948)
★ Multilayer Feedforward NN Without Deep Learning (Rosenblatt, 1958)
★ First Published Learning RNNs (Amari and others, ~1972)
★ First Deep Learning (Ivakhnenko & Lapa, 1965)
★ Deep Learning by Stochastic Gradient Descent (Amari, 1967-68)
★ ReLUs (Fukushima, 1969)
★ Backpropagation (Linnainmaa, 1970); precursor (Kelley, 1960)
★ Backpropagation for NNs (Werbos, 1982)
★ First Deep Convolutional NN (Fukushima, 1979); later combined with Backprop (Waibel 1987, Zhang 1988).
★ Metalearning or Learning to Learn (Schmidhuber, 1987)
★ Generative Adversarial Networks / Artificial Curiosity / NN Online Planners (Schmidhuber, Feb 1990; see the G in Generative AI and ChatGPT)
★ NNs Learn to Generate Subgoals and Work on Command (Schmidhuber, April 1990)
★ NNs Learn to Program NNs: Unnormalized Linear Transformer (Schmidhuber, March 1991; see the T in ChatGPT)
★ Deep Learning by Self-Supervised Pre-Training. Distilling NNs (Schmidhuber, April 1991; see the P in ChatGPT)
★ Experiments with Pre-Training; Analysis of Vanishing/Exploding Gradients, Roots of Long Short-Term Memory / Highway Nets / ResNets (Hochreiter, June 1991, further developed 1999-2015 with other students of Schmidhuber)
★ LSTM journal paper (1997, most cited AI paper of the 20th century)
★ xLSTM (Hochreiter, 2024)
★ Reinforcement Learning Prompt Engineer for Abstract Reasoning and Planning (Schmidhuber 2015)
★ Mindstorms in Natural Language-Based Societies of Mind (2023 paper by Schmidhuber's team)
arxiv.org/abs/2305.17066
★ Bremermann's physical limit of computation (1982)
EXTERNAL LINKS
CogX 2018 - Professor Juergen Schmidhuber
youtube.com/watch?v=17shdT9-wuA
Discovering Neural Nets with Low Kolmogorov Complexity and High Generalization Capability (Neural Networks, 1997)
https://sferics.idsia.ch/pub/juergen/loconet.pdf
The paradox at the heart of mathematics: Gödel's Incompleteness Theorem - Marcus du Sautoy
youtube.com/watch?v=I4pQbo5MQOs
The Philosophy of Science - Hilary Putnam & Bryan Magee (1977)
youtube.com/watch?v=JJB2q8ufAgk
Optimal Ordered Problem Solver
arxiv.org/abs/cs/0207097
Levin's Universal Search from 1973
rjlipton.com/2011/03/14/levins-great-discoveries
https://people.idsia.ch/~juergen/optimalsearch.html
On Learning to Think
arxiv.org/abs/1511.09249
Mindstorms in Natural Language-Based Societies of Mind
Untersuchungen zu dynamischen neuronalen Netzen
https://www.bioinf.jku.at/publications/older/3804.pdf
Evolutionary Principles in Self-Referential Learning
https://people.idsia.ch/~juergen/diploma1987ocr.pdf
Hans-Joachim Bremermann
en.wikipedia.org/wiki/Bremermann%27s_limit
Highway Networks
arxiv.org/abs/1505.00387
https://people.idsia.ch/~juergen/highway-networks.html
The principles of Deep Learning Theory
amzn.to/3WJtPaj
Understanding Deep Learning amzn.to/4doDk63
Discovering Problem Solutions with Low Kolmogorov Complexity and High Generalization Capability (ICML 1995)
https://sferics.idsia.ch/pub/juergen/icmlkolmogorov.pdf
"History of Modern AI and Deep Learning":
https://people.idsia.ch/~juergen/deep-learning-history.html
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmented generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Show notes:
* Domingos' views on AI regulation and why he believes it's misguided
* His thoughts on the current state of AI technology and its limitations
* Discussion of his novel "2040", a satirical take on AI and tech culture
* Explanation of his work on "tensor logic", which aims to unify neural networks and symbolic AI
* Critiques of other approaches in AI, including those of OpenAI and Gary Marcus
* Thoughts on the AI "bubble" and potential future developments in the field
Prof. Pedro Domingos:
https://x.com/pmddomingos
2040: A Silicon Valley Satire [Pedro's new book]
amzn.to/3T51ISd
Pod version: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/AI-should-NOT-be-regulated-at-all----Prof--Pedro-Domingos-e2njada
TOC:
00:00:00 Intro
00:06:31 Bio
00:08:40 Filmmaking skit
00:10:35 AI and the wisdom of crowds
00:19:49 Social Media
00:27:48 Master algorithm
00:30:48 Neurosymbolic AI / abstraction
00:39:01 Language
00:45:38 Chomsky
01:00:49 2040 Book
01:18:03 Satire as a shield for criticism?
01:29:12 AI Regulation / Gary Marcus
01:52:37 Copyright
01:56:11 Stochastic parrots come home to roost
02:00:03 Privacy
02:01:55 LLM ecosystem
02:05:06 Tensor logic
Refs:
The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World [Pedro Domingos]
amzn.to/3MiWs9B
Rebooting AI: Building Artificial Intelligence We Can Trust [Gary Marcus]
amzn.to/3AAywvL
Flash Boys [Michael Lewis]
amzn.to/4dUGm1M
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api
This is video 4/13 from ICML 2024
Andrew's site:
andrewilyas.com
https://x.com/andrew_ilyas
TOC:
00:00:00 - Introduction and Andrew's background
00:03:52 - Overview of the machine learning pipeline
00:06:31 - Data modeling paper discussion
00:26:28 - TRAK: Evolution of data modeling work
00:43:58 - Discussion on abstraction, reasoning, and neural networks
00:53:16 - "Adversarial Examples Are Not Bugs, They Are Features" paper
01:03:24 - Types of features learned by neural networks
01:10:51 - Black box attacks paper
01:15:39 - Work on data collection and bias
01:25:48 - Future research plans and closing thoughts
Pod version: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Adversarial-Examples-and-Data-Modelling---Andrew-Ilyas-MIT-e2nfov9
References:
Adversarial Examples Are Not Bugs, They Are Features
arxiv.org/pdf/1905.02175
TRAK: Attributing Model Behavior at Scale
arxiv.org/pdf/2303.14186
Datamodels: Predicting Predictions from Training Data
arxiv.org/pdf/2202.00622
Adversarial Examples Are Not Bugs, They Are Features
arxiv.org/pdf/1905.02175
IMAGENET-TRAINED CNNS
arxiv.org/pdf/1811.12231
ZOO: Zeroth Order Optimization Based Black-box
arxiv.org/pdf/1708.03999
A Spline Theory of Deep Networks
https://proceedings.mlr.press/v80/balestriero18b/balestriero18b.pdf
Scaling Monosemanticity
https://transformer-circuits.pub/2024/scaling-monosemanticity/
Adversarial Examples Are Not Bugs, They Are Features
gradientscience.org/adv
Adversarial Robustness Limits via Scaling-Law and Human-Alignment Studies
https://proceedings.mlr.press/v235/bartoldson24a.html
Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors
arxiv.org/abs/1807.07978
Estimation of Standard Auction Models
arxiv.org/abs/2205.02060
From ImageNet to Image Classification: Contextualizing Progress on Benchmarks
arxiv.org/abs/2005.11295
Estimation of Standard Auction Models
arxiv.org/abs/2205.02060
What Makes A Good Fisherman? Linear Regression under Self-Selection Bias
arxiv.org/abs/2205.03246
Towards Tracing Factual Knowledge in Language Models Back to the
Training Data [Akyürek]
arxiv.org/pdf/2205.11482
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Joscha takes us on a tour de force through history, philosophy, and cutting-edge computer science, teasing us to rethink what we know about minds, machines, and the world around us. Joscha believes we should blur the lines between human, artificial, and natural intelligence, and argues that consciousness might be more widespread and interconnected than we ever thought possible.
Dr. Joscha Bach
https://x.com/Plinz
This is video 2/9 from our coverage of AGI-24 in Seattle agi-conf.org/2024
Watch the official MLST interview with Joscha which we did right after this talk on our Patreon now on early access - patreon.com/posts/joscha-bach-110199676 (you also get access to our private discord and biweekly calls)
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Joscha-Bach---AGI24-Keynote-Cyberanimism-e2neeui
TOC:
00:00:00 Introduction: AGI and Cyberanimism
00:03:57 The Nature of Consciousness
00:08:46 Aristotle's Concepts of Mind and Consciousness
00:13:23 The Hard Problem of Consciousness
00:16:17 Functional Definition of Consciousness
00:20:24 Comparing LLMs and Human Consciousness
00:26:52 Testing for Consciousness in AI Systems
00:30:00 Animism and Software Agents in Nature
00:37:02 Plant Consciousness and Ecosystem Intelligence
00:40:36 The California Institute for Machine Consciousness
00:44:52 Ethics of Conscious AI and Suffering
00:46:29 Philosophical Perspectives on Consciousness
00:49:55 Q&A: Formalisms for Conscious Systems
00:53:27 Coherence, Self-Organization, and Compute Resources
Refs:
Aristotle's work on the soul and consciousness
Richard Dawkins' work on genes and evolution
Gerald Edelman's concept of Neural Darwinism
Thomas Metzinger's book "Being No One"
Yoshua Bengio's concept of the "consciousness prior"
Stuart Hameroff's theories on microtubules and consciousness
Christof Koch's work on consciousness
Daniel Dennett's "Cartesian Theater" concept
Giulio Tononi's Integrated Information Theory
Mike Levin's work on organismal intelligence
The concept of animism in various cultures
Freud's model of the mind
Buddhist perspectives on consciousness and meditation
The Genesis creation narrative (for its metaphorical interpretation)
California Institute for Machine Consciousness
This is part 1, we will be releasing an in-depth interview with Gary in the coming weeks.
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Gary Marcus criticized current large language models (LLMs) and generative AI for their unreliability, tendency to hallucinate, and inability to truly understand concepts.
Marcus argued that the AI field is experiencing diminishing returns with current approaches, particularly the "scaling hypothesis" that simply adding more data and compute will lead to AGI.
He advocated for a hybrid approach to AI that combines deep learning with symbolic AI, emphasizing the need for systems with deeper conceptual understanding.
Marcus highlighted the importance of developing AI with innate understanding of concepts like space, time, and causality.
He expressed concern about the moral decline in Silicon Valley and the rush to deploy potentially harmful AI technologies without adequate safeguards.
Marcus predicted a possible upcoming "AI winter" due to inflated valuations, lack of profitability, and overhyped promises in the industry.
He stressed the need for better regulation of AI, including transparency in training data, full disclosure of testing, and independent auditing of AI systems.
Marcus proposed the creation of national and global AI agencies to oversee the development and deployment of AI technologies.
He concluded by emphasizing the importance of interdisciplinary collaboration, focusing on robust AI with deep understanding, and implementing smart, agile governance for AI and AGI.
Pre-order Gary's new book here:
Taming Silicon Valley: How We Can Ensure That AI Works for Us
amzn.to/4fO46pY
Filmed at the AGI-24 conference:
agi-conf.org/2024
Refs:
Closed source vs open-source models slide ~24 mins
Fine-tune Llama 3.1 Ultra-Efficiently with Unsloth (Maxime Labonne/Liquid AI)
huggingface.co/blog/mlabonne/sft-llama3
TOC:
00:00:00 Introduction
00:02:34 Introduction by Ben G
00:05:17 Gary Marcus begins talk
00:07:38 Critiquing current state of AI
00:12:21 Lack of progress on key AI challenges
00:16:05 Continued reliability issues with AI
00:19:54 Economic challenges for AI industry
00:25:11 Need for hybrid AI approaches
00:29:58 Moral decline in Silicon Valley
00:34:59 Risks of current generative AI
00:40:43 Need for AI regulation and governance
00:49:21 Concluding thoughts
00:54:38 Q&A: Cycles of AI hype and winters
01:00:10 Predicting a potential AI winter
01:02:46 Discussion on interdisciplinary approach
01:05:46 Question on regulating AI
01:07:27 Ben G's perspective on AI winter
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Key points covered include:
A method for describing transformer predictions using n-gram statistics without relying on internal mechanisms.
The discovery of a technique to detect overfitting in large language models without using holdout sets.
Observations on curriculum learning, showing how transformers progress from simpler to more complex rules during training.
Discussion of distance measures used in the analysis, particularly the variational distance.
Exploration of model sizes, training dynamics, and their impact on the results.
We also touch on philosophical aspects of describing versus explaining AI behavior, and the challenges in understanding the abstractions formed by neural networks. Nguyen concludes by discussing potential future research directions, including attempts to convert descriptions of transformer behavior into explanations of internal mechanisms.
Timothy Nguyen's earned his B.S. and Ph.D. in mathematics from Caltech and MIT, respectively. He held positions as Research Assistant Professor at the Simons Center for Geometry and Physics (2011-2014) and Visiting Assistant Professor at Michigan State University (2014-2017). During this time, his research expanded into high-energy physics, focusing on mathematical problems in quantum field theory. His work notably provided a simplified and corrected formulation of perturbative path integrals.
Since 2017, Nguyen has been working in industry, applying his expertise to machine learning. He is currently at DeepMind, where he contributes to both fundamental research and practical applications of deep learning to solve real-world problems.
Refs:
The Cartesian Cafe
youtube.com/@TimothyNguyen
Understanding Transformers via N-Gram Statistics
researchgate.net/publication/382204056_Understanding_Transformers_via_N-Gram_Statistics
TOC
00:00:00 Timothy Nguyen's background
00:02:50 Paper overview: transformers and n-gram statistics
00:04:55 Template matching and hash table approach
00:08:55 Comparing templates to transformer predictions
00:12:01 Describing vs explaining transformer behavior
00:15:36 Detecting overfitting without holdout sets
00:22:47 Curriculum learning in training
00:26:32 Distance measures in analysis
00:28:58 Model sizes and training dynamics
00:30:39 Future research directions
00:32:06 Conclusion and future topics
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Cohere Command R model series: cohere.com/command
Jay Alamaar:
https://x.com/jayalammar
Buy Jay's new book here!
Hands-On Large Language Models: Language Understanding and Generation
amzn.to/4fzOUgh
TOC:
00:00:00 Introduction to Jay Alammar and AI Education
00:01:47 Cohere's Approach to RAG and AI Re-ranking
00:07:15 Implementing AI in Enterprise: Challenges and Solutions
00:09:26 Jay's Role at Cohere and the Importance of Learning in Public
00:15:16 The Evolution of AI in Industry: From Deep Learning to LLMs
00:26:12 Expert Advice for Newcomers in Machine Learning
00:32:39 The Power of Semantic Search and Embeddings in AI Systems
00:37:59 Jay Alammar's Journey as an AI Educator and Visualizer
00:43:36 Visual Learning in AI: Making Complex Concepts Accessible
00:47:38 Strategies for Keeping Up with Rapid AI Advancements
00:49:12 The Future of Transformer Models and AI Architectures
00:51:40 Evolution of the Transformer: From 2017 to Present
00:54:19 Preview of Jay's Upcoming Book on Large Language Models
Disclaimer: This is the fourth video from our Cohere partnership. We were not told what to say in the interview, and didn't edit anything out from the interview. Note also that this combines several previously unpublished interviews from Jay into one, the earlier one at Tim's house was shot in Aug 2023, and the more recent one in Toronto in May 2024.
Refs:
The Illustrated Transformer
jalammar.github.io/illustrated-transformer
Attention Is All You Need
arxiv.org/abs/1706.03762
The Unreasonable Effectiveness of Recurrent Neural Networks
http://karpathy.github.io/2015/05/21/rnn-effectiveness
Neural Networks in 11 Lines of Code
iamtrask.github.io/2015/07/12/basic-python-network
Understanding LSTM Networks (Chris Olah's blog post)
http://colah.github.io/posts/2015-08-Understanding-LSTMs
Luis Serrano's YouTube Channel
youtube.com/channel/UCgBncpylJ1kiVaPyP-PZauQ
Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
arxiv.org/abs/1908.10084
GPT (Generative Pre-trained Transformer) models
jalammar.github.io/illustrated-gpt2
openai.com/research/gpt-4
BERT (Bidirectional Encoder Representations from Transformers)
jalammar.github.io/illustrated-bert
arxiv.org/abs/1810.04805
RoPE (Rotary Positional Encoding)
arxiv.org/abs/2104.09864 (Linked paper discussing rotary embeddings)
Grouped Query Attention
arxiv.org/pdf/2305.13245
RLHF (Reinforcement Learning from Human Feedback)
openai.com/research/learning-from-human-preferences
arxiv.org/abs/1706.03741
DPO (Direct Preference Optimization)
arxiv.org/abs/2305.18290
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
Daniel Cahn (who is also hiring ML engineers by the way!)
https://x.com/thecahnartist?lang=en
linkedin.com/in/cahnd
thinkingmachinespodcast.com
TOC:
00:00:00 Intro
00:01:56 Therapy effectiveness vs drugs and societal implications
00:04:02 Mental health categories: Iatrogenesis and social constructs
00:10:19 Psychiatric treatment models and cognitive perspectives
00:13:30 AI design and human-like interactions: Intentionality debates
00:20:04 AI in therapy: Ethics, anthropomorphism, and loneliness mitigation
00:28:13 Therapy efficacy: Neuroplasticity, suffering, and AI placebos
00:33:29 AI's impact on human agency and cognitive modeling
00:41:17 Social media's effects on brain structure and behavior
00:50:46 AI ethics: Altering values and free will considerations
01:00:00 Work value perception and personal identity formation
01:13:37 Free will, agency, and mutable personal identity in therapy
01:24:27 AI in healthcare: Challenges, ethics, and therapy improvements
01:53:25 AI development: Societal impacts and cultural implications
Pod version: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Can-AI-therapy-be-more-effective-than-drugs-e2mvhd0
Refs:
Bad Therapy: Why the Kids Aren't Growing Up (Abigail Shrier)
amzn.to/3WSLga4
Irreversible Damage: Teenage Girls and the Transgender Craze (Shrier)
amzn.to/4d9kpfx (note that this book will be offensive to trans folks, forewarned - Tim hasn't read it/isn't taking a position)
The Coddling of the American Mind (Haidt)
amzn.to/3WMtPaS
The Anxious Generation (Haidt)
amzn.to/3SB9MJU
Lost Connections (Johan Hari)
amzn.to/3SC7471
(Note the book cover of "Hillbilly Elegy" was a mistake, I meant to link to "The Minds of Billy Milligan") i.e. a book about multiple personality disorder
amzn.to/4dUHZg7
Noise (Kahneman)
amzn.to/4ftMgsi
Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR)
psychiatry.org/psychiatrists/practice/dsm
Biopsychosocial model
en.wikipedia.org/wiki/Biopsychosocial_model
The serotonin theory of depression: how the media got it all wrong
pharmaceutical-journal.com/article/feature/the-serotonin-theory-of-depression-how-the-media-got-it-all-wrong
The ELIZA effect
en.wikipedia.org/wiki/ELIZA_effect
How to Win Friends and Influence People
amazon.co.uk/How-Win-Friends-Influence-People/dp/0091906814
Evidence of Human-Level Bonds Established With a Digital Conversational Agent: Cross-sectional, Retrospective Observational Study (JMIR)
formative.jmir.org/2021/5/e27868
COMPUTING MACHINERY AND INTELLIGENCE (Turing - imitation game)
cs.ox.ac.uk/activities/ieg/e-library/sources/t_article.pdf
Decision transformers
arxiv.org/abs/2106.01345
Loneliness and suicide mitigation for students using GPT3-enabled chatbots
nature.com/articles/s44184-023-00047-6
Little Treatments, Big Effects (Jessica Schleider)
amzn.to/4deN7vh
Why Greatness Cannot Be Planned (Stanley and Lehman) [Possibly the best book ever written]
amzn.to/3yxykNf
Machine Love (Lehman)
arxiv.org/abs/2302.09248
Infinite games
amzn.to/3LYF3Tq
OpenAI model spec
openai.com/index/introducing-the-model-spec
A Brief History of Intelligence (Bennett - interview dropping soon)
amzn.to/3WTWs5Z
Bullshit Jobs: A Theory
amzn.to/46ENSeO
John Carmack take - https://x.com/ID_AA_Carmack/status/1796622337963389412
Determined (Robert M Sapolsky)
amzn.to/4dc8KN3
So You've Been Publicly Shamed (Jon Ronson)
amzn.to/3WDLzE9
The Examined Life (Stephen Grosz)
amzn.to/3SF6s0H
Thinking, Fast and Slow (Daniel Kahneman)
amzn.to/3WzEZyx
Decision Transformer: Reinforcement Learning via Sequence Modeling
arxiv.org/pdf/2106.01345
Ender's Game
amzn.to/3yxXNGq
Magic Pill: The Extraordinary Benefits and Disturbing Risks of the New Weight Loss Drugs
amzn.to/4ccrc6Q
Thinking Machines (Daniel’s podcast!)
thinkingmachinespodcast.com
The Upside of Stress (Daniel referenced this on a part of the conversation which we edited out)
amzn.to/3LYMTww
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
This is 2/13 of our #ICML2024 series
TOC
[00:00:00] Intro
[00:02:06] Bio
[00:03:02] LLMs are n-gram models on steroids
[00:07:26] Is natural language a formal language?
[00:08:34] Natural language is formal?
[00:11:01] Do LLMs reason?
[00:19:13] Definition of reasoning
[00:31:40] Creativity in reasoning
[00:50:27] Chollet's ARC challenge
[01:01:31] Can we reason without verification?
[01:10:00] LLMs cant solve some tasks
[01:19:07] LLM Modulo framework
[01:29:26] Future trends of architecture
[01:34:48] Future research directions
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Prof--Subbarao-Kambhampati---LLMs-dont-reason--they-memorize-ICML2024-213-e2mjcse
Subbarao Kambhampati:
https://x.com/rao2z
Interviewer: Dr. Tim Scarfe
Refs:
Can LLMs Really Reason and Plan?
cacm.acm.org/blogcacm/can-llms-really-reason-and-plan
On the Planning Abilities of Large Language Models : A Critical Investigation
arxiv.org/pdf/2305.15771
Chain of Thoughtlessness? An Analysis of CoT in Planning
arxiv.org/pdf/2405.04776
On the Self-Verification Limitations of Large Language Models on Reasoning and Planning Tasks
arxiv.org/pdf/2402.08115
LLMs Can't Plan, But Can Help Planning in LLM-Modulo Frameworks
arxiv.org/pdf/2402.01817
Embers of Autoregression: Understanding Large Language
Models Through the Problem They are Trained to Solve
arxiv.org/pdf/2309.13638
arxiv.org/abs/2402.04210
"Task Success" is not Enough
Faith and Fate: Limits of Transformers on Compositionality "finetuning multiplication with four digit numbers" (added after pub)
arxiv.org/pdf/2305.18654
Partition function (number theory) (Srinivasa Ramanujan and G.H. Hardy's work)
en.wikipedia.org/wiki/Partition_function_(number_theory)
Poincaré conjecture
en.wikipedia.org/wiki/Poincar%C3%A9_conjecture
Gödel's incompleteness theorems
en.wikipedia.org/wiki/G%C3%B6del%27s_incompleteness_theorems
ROT13 (Rotate13, "rotate by 13 places")
en.wikipedia.org/wiki/ROT13
A Mathematical Theory of Communication (C. E. SHANNON)
https://people.math.harvard.edu/~ctm/home/text/others/shannon/entropy/entropy.pdf
Sparks of AGI
arxiv.org/abs/2303.12712
Kambhampati thesis on speech recognition (1983)
https://rakaposhi.eas.asu.edu/rao-btech-thesis.pdf
PlanBench: An Extensible Benchmark for Evaluating Large Language Models on Planning and Reasoning about Change
arxiv.org/abs/2206.10498
Explainable human-AI interaction
link.springer.com/book/10.1007/978-3-031-03767-2
Tree of Thoughts
arxiv.org/abs/2305.10601
On the Measure of Intelligence (ARC Challenge)
arxiv.org/abs/1911.01547
Getting 50% (SoTA) on ARC-AGI with GPT-4o (Ryan Greenblatt ARC solution)
redwoodresearch.substack.com/p/getting-50-sota-on-arc-agi-with-gpt
PROGRAMS WITH COMMON SENSE (John McCarthy) - "AI should be an advice taker program"
https://www.cs.cornell.edu/selman/cs672/readings/mccarthy-upd.pdf
Original chain of thought paper
arxiv.org/abs/2201.11903
ICAPS 2024 Keynote: Dale Schuurmans on "Computing and Planning with Large Generative Models" (COT)
youtube.com/watch?v=YnMqbpdHcaY
The Hardware Lottery (Hooker)
arxiv.org/abs/2009.06489
A Path Towards Autonomous Machine Intelligence (JEPA/LeCun)
openreview.net/pdf?id=BZ5a1r-kVsf
AlphaGeometry
nature.com/articles/s41586-023-06747-5
FunSearch
nature.com/articles/s41586-023-06924-6
Emergent Abilities of Large Language Models
arxiv.org/abs/2206.07682
Language models are not naysayers (Negation in LLMs)
arxiv.org/abs/2306.08189
The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
arxiv.org/abs/2309.12288
Embracing negative results
openreview.net/forum?id=3RXAiU7sss
MLST is sponsored by Brave:
The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmentated generation. Try it now - get 2,000 free queries monthly at http://brave.com/api.
This is part of our ICML 2024 series (1/13)
Sayash Kapoor is a computer science Ph.D. candidate at Princeton University's Center for Information Technology Policy. His research focuses on the societal impact of AI. Kapoor has previously worked on AI in both industry and academia, with experience at Facebook, Columbia University, and EPFL Switzerland. He is a recipient of a best paper award at ACM FAccT and an impact recognition award at ACM CSCW. Notably, Kapoor was included in TIME's inaugural list of the 100 most influential people in AI.
Sayash Kapoor
https://x.com/sayashk
https://www.cs.princeton.edu/~sayashk/
Arvind Narayanan (other half of the AI Snake Oil duo)
https://x.com/random_walker
AI existential risk probabilities are too unreliable to inform policy
aisnakeoil.com/p/ai-existential-risk-probabilities
Pre-order AI Snake Oil Book
amzn.to/4fq2HGb
AI Snake Oil blog
aisnakeoil.com
AI Agents That Matter
arxiv.org/abs/2407.01502
Shortcut learning in deep neural networks
semanticscholar.org/paper/Shortcut-learning-in-deep-neural-networks-Geirhos-Jacobsen/1b04936c2599e59b120f743fbb30df2eed3fd782
77% Of Employees Report AI Has Increased Workloads And Hampered Productivity, Study Finds
forbes.com/sites/bryanrobinson/2024/07/23/employees-report-ai-increased-workload
Pod version: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Sayash-Kapoor---How-seriously-should-we-take-AI-X-risk--ICML-113-e2mhuoc
TOC:
00:00:00 Intro
00:01:57 How seriously should we take Xrisk threat?
00:02:55 Risk too unrealiable to inform policy
00:10:20 Overinflated risks
00:12:05 Perils of utility maximisation
00:13:55 Scaling vs airplane speeds
00:17:31 Shift to smaller models?
00:19:08 Commercial LLM ecosystem
00:22:10 Synthetic data
00:24:09 Is AI complexifying our jobs?
00:25:50 Does ChatGPT make us dumber or smarter?
00:26:55 Are AI Agents overhyped?
00:28:12 Simple vs complex baselines
00:30:00 Cost tradeoff in agent design
00:32:30 Model eval vs downstream perf
00:36:49 Shortcuts in metrics
00:40:09 Standardisation of agent evals
00:41:21 Humans in the loop
00:43:54 Levels of agent generality
00:47:25 ARC challenge
We explore why this approach, recently adopted in both US and EU AI policies, may be problematic and oversimplified. Sara explains the limitations of using raw computational power as a measure of AI capability or risk, and discusses the complex relationship between compute, data, and model architecture.
Equally important, we go into Sara's work on "The AI Language Gap." This research highlights the challenges and inequalities in developing AI systems that work across multiple languages. Sara discusses how current AI models, predominantly trained on English and a handful of high-resource languages, fail to serve the linguistic diversity of our global population. We explore the technical, ethical, and societal implications of this gap, and discuss potential solutions for creating more inclusive and representative AI systems.
We broadly discuss the relationship between language, culture, and AI capabilities, as well as the ethical considerations in AI development and deployment.
Pod version: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Sara-Hooker---Why-US-AI-Act-Compute-Thresholds-Are-Misguided-e2m6qm4
TOC:
[00:00:00] Intro
[00:02:12] FLOPS paper
[00:26:42] Hardware lottery
[00:30:22] The Language gap
[00:33:25] Safety
[00:38:31] Emergent
[00:41:23] Creativity
[00:43:40] Long tail
[00:44:26] LLMs and society
[00:45:36] Model bias
[00:48:51] Language and capabilities
[00:52:27] Ethical frameworks and RLHF
Sara Hooker
sarahooker.me
linkedin.com/in/sararosehooker
scholar.google.com/citations?user=2xy6h3sAAAAJ&hl=en
https://x.com/sarahookr
Interviewer: Tim Scarfe
Refs
The AI Language gap
cohere.com/research/papers/the-AI-language-gap.pdf
On the Limitations of Compute Thresholds as a Governance Strategy.
arxiv.org/pdf/2407.05694v1
The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce Harm
arxiv.org/pdf/2406.18682
Cohere Aya
cohere.com/research/aya
RLHF Can Speak Many Languages: Unlocking Multilingual Preference Optimization for LLMs
arxiv.org/pdf/2407.02552
Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs
arxiv.org/pdf/2402.14740
Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence
whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence
EU AI Act
https://www.europarl.europa.eu/doceo/document/TA-9-2024-0138_EN.pdf
The bitter lesson
http://www.incompleteideas.net/IncIdeas/BitterLesson.html
Neel Nanda interview
youtube.com/watch?v=_Ygf0GnlwmY
Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet
https://transformer-circuits.pub/2024/scaling-monosemanticity/
Chollet's ARC challenge
github.com/fchollet/ARC-AGI
Ryan Greenblatt on ARC
youtube.com/watch?v=z9j3wB1RRGA
Disclaimer: This is the third video from our Cohere partnership. We were not told what to say in the interview, and didn't edit anything out from the interview.
We explore the dangers of anthropomorphizing AI, the limitations of current language in describing AI capabilities, and the fascinating intersection of philosophy and artificial intelligence.
Show notes and full references: docs.google.com/document/d/1ICtBI574W-xGi8Z2ZtUNeKWiOiGZ_DRsp9EnyYAISws/edit?usp=sharing
Prof Murray Shanahan:
doc.ic.ac.uk/~mpsha (look at his selected publications)
scholar.google.co.uk/citations?user=00bnGpAAAAAJ&hl=en
en.wikipedia.org/wiki/Murray_Shanahan
https://x.com/mpshanahan
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Prof--Murray-Shanahan---Machines-Dont-Think-Like-Us-e2m1499
Interviewer: Dr. Tim Scarfe
Refs (links in the Google doc linked above):
Role play with large language models
Waluigi effect
"Conscious Exotica" - Paper by Murray Shanahan (2016)
"Simulators" - Article by Janis from LessWrong
"Embodiment and the Inner Life" - Book by Murray Shanahan (2010)
"The Technological Singularity" - Book by Murray Shanahan (2015)
"Simulacra as Conscious Exotica" - Paper by Murray Shanahan (newer paper of the original focussed on LLMs)
A recent paper by Anthropic on using autoencoders to find features in language models (referring to the "Scaling Monosemanticity" paper)
Work by Peter Godfrey-Smith on octopus consciousness
"Metaphors We Live By" - Book by George Lakoff (1980s)
Work by Aaron Sloman on the concept of "space of possible minds" (1984 article mentioned)
Wittgenstein's "Philosophical Investigations" (posthumously published)
Daniel Dennett's work on the "intentional stance"
Alan Turing's original paper on the Turing Test (1950)
Thomas Nagel's paper "What is it like to be a bat?" (1974)
John Searle's Chinese Room Argument (mentioned but not detailed)
Work by Richard Evans on tackling reasoning problems
Claude Shannon's quote on knowledge and control
"Are We Bodies or Souls?" - Book by Richard Swinburne
Reference to work by Ethan Perez and others at Anthropic on potential deceptive behavior in language models
Reference to a paper by Murray Shanahan and Antonia Creswell on the "selection inference framework"
Mention of work by Francois Chollet, particularly the ARC (Abstraction and Reasoning Corpus) challenge
Reference to Elizabeth Spelke's work on core knowledge in infants
Mention of Karl Friston's work on planning as inference (active inference)
The film "Ex Machina" - Murray Shanahan was the scientific advisor
"The Waluigi Effect"
Anthropic's constitutional AI approach
Loom system by Lara Reynolds and Kyle McDonald for visualizing conversation trees
DeepMind's AlphaGo (mentioned multiple times as an example)
Mention of the "Golden Gate Claude" experiment
Reference to an interview Tim Scarfe conducted with University of Toronto students about self-attention controllability theorem
Mention of an interview with Irina Rish
Reference to an interview Tim Scarfe conducted with Daniel Dennett
Reference to an interview with Maria Santa Caterina
Mention of an interview with Philip Goff
Nick Chater and Martin Christianson's book ("The Language Game: How Improvisation Created Language and Changed the World")
Peter Singer's work from 1975 on ascribing moral status to conscious beings
Demis Hassabis' discussion on the "ladder of creativity"
Reference to B.F. Skinner and behaviorism
TOC:
00:00:00 Intro
00:05:49 Simulators
00:11:04 The 20 questions game and simulacra stickiness
00:18:50 Murray's experience with Claude 3
00:30:04 RLHF
00:32:41 Anthropic Golden Gate Bridge
00:37:05 Agency
00:41:05 Embodiment and knowledge acquisition
00:57:51 ARC
01:03:31 The conscious stance
01:13:58 Space of possible minds
01:11:05 part 2: Wittgenstein private language argument / subjectivity
01:29:58 Conscious exotica
01:33:23 Dennett and intentional stance
01:40:58 Anthropomorphisation
01:46:47 Reasoning
01:53:56 Turing test
02:04:41 Nagel's bat
02:08:08 Mark Bishop and Idealism/CRA
02:09:32 Panpsychism
Sponsor:
Sign up to Kalshi here kalshi.onelink.me/1r91/mlst -- the first 500 traders who deposit $100 will get a free $20 credit! Important disclaimer - In case it's not obvious - this is basically gambling and a *high risk* activity - only trade what you can afford to lose.
We discuss:
- Ryan's unique approach to solving the ARC Challenge and achieving impressive results.
- The strengths and weaknesses of current AI models.
- How AI and humans differ in learning and reasoning.
- Combining various techniques to create smarter AI systems.
- The potential risks and future advancements in AI, including the idea of agentic AI.
https://x.com/RyanPGreenblatt
redwoodresearch.org
TOC
00:00:00 Intro
00:01:38 Prelude on goals in LLMs
00:02:42 Ryan intro
00:03:11 Ryan's ARC Challenge Approach
00:38:15 Language models, reasoning and agency
01:14:14 Timelines on superintelligence
01:27:05 Growth of superintelligence
02:06:41 Reflections on ARC
02:11:49 Why wouldn't AI knowledge be subjective
Host: Dr. Tim Scarfe
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Ryan-Greenblatt---Solving-ARC-with-GPT4o-e2lnplq
Refs:
Getting 50% (SoTA) on ARC-AGI with GPT-4o [Ryan Greenblatt]
redwoodresearch.substack.com/p/getting-50-sota-on-arc-agi-with-gpt
On the Measure of Intelligence [Chollet]
arxiv.org/abs/1911.01547
Connectionism and Cognitive Architecture: A Critical Analysis [Jerry A. Fodor and Zenon W. Pylyshyn]
https://ruccs.rutgers.edu/images/personal-zenon-pylyshyn/proseminars/Proseminar13/ConnectionistArchitecture.pdf
Software 2.0 [Andrej Karpathy]
karpathy.medium.com/software-2-0-a64152b37c35
Why Greatness Cannot Be Planned: The Myth of the Objective [Kenneth Stanley]
amzn.to/3Wfy2E0
Biographical account of Terence Tao’s mathematical development. [M.A.(KEN) CLEMENTS]
gwern.net/doc/iq/high/smpy/1984-clements.pdf
Model Evaluation and Threat Research (METR)
metr.org
Why Tool AIs Want to Be Agent AIs
gwern.net/tool-ai
Simulators - Janus
lesswrong.com/posts/vJFdjigzmcXMhNTsx/simulators
AI Control: Improving Safety Despite Intentional Subversion
lesswrong.com/posts/d9FJHawgkiMSPjagR/ai-control-improving-safety-despite-intentional-subversion
arxiv.org/abs/2312.06942
What a Compute-Centric Framework Says About Takeoff Speeds
openphilanthropy.org/research/what-a-compute-centric-framework-says-about-takeoff-speeds
Global GDP over the long run
ourworldindata.org/grapher/global-gdp-over-the-long-run?yScale=log
Safety Cases: How to Justify the Safety of Advanced AI Systems
arxiv.org/abs/2403.10462
The Danger of a “Safety Case"
http://sunnyday.mit.edu/The-Danger-of-a-Safety-Case.pdf
The Future Of Work Looks Like A UPS Truck (~02:15:50)
npr.org/sections/money/2014/05/02/308640135/episode-536-the-future-of-work-looks-like-a-ups-truck
SWE-bench
swebench.com
Using DeepSpeed and Megatron to Train Megatron-Turing NLG
530B, A Large-Scale Generative Language Model
arxiv.org/pdf/2201.11990
Algorithmic Progress in Language Models
epochai.org/blog/algorithmic-progress-in-language-models
Aidan shares his personal insights into the world of AI and LLMs and Cohere's unique approach to solving real-world business problems, and how their models are set apart from the competition. Aidan reveals how they are making major strides in AI technology, discussing everything from last mile customer engineering to the robustness of prompts and future architectures.
He also touches on the broader implications of AI for society, including potential risks and the role of regulation. He discusses Cohere's guiding principles and the health the of startup scene. With a particular focus on enterprise applications. Aidan provides a rare look into the internal workings of Cohere and their vision for driving productivity and innovation.
cohere.com
https://x.com/aidangomez
Check out Cohere's amazing new Command R* models here
cohere.com/command
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Aiden-Gomez---CEO-of-Cohere-AIs-Inner-Monologue--Crucial-for-Reasoning-e2lf8fg
TOC:
00:00:00 Intro
00:01:48 Guiding principles of Cohere
00:02:31 Last mile / customer engineering
00:04:25 Prompt brittleness
00:06:14 Robustness and "delving"
00:10:12 Command R models and catch up
00:12:32 Are LLMs saturating / specialisation
00:16:11 Intelligence
00:21:28 Predictive architectures, data vs inductive priors
00:25:55 Agentic systems
00:28:11 Differentiation
00:33:35 X-Risk / Bostrom
00:39:30 Changing relationship with technology
00:45:08 Policy
00:49:01 Startup scene
00:52:44 Biggest mistake?
00:53:50 Management style
00:56:38 Culture in different Cohere offices?
Disclaimer: This is the second video from our Cohere partnership. We were not told what to say in the interview, and didn't edit anything out from the interview.
Philip discuss another idea called cosmopsychism, which is a theory that suggests the entire universe is a single conscious entity. Instead of individual minds (like human minds) being separate and independent, they are seen as parts of the universe's larger, unified consciousness. In simpler terms, it means that the universe itself has a mind, and our individual consciousnesses are just small parts of this greater, universal mind.
Philip thinks science can't fully explain what it's like to feel things, like the color red or the taste of chocolate. He says we need to include consciousness in our science to understand everything better.
We edited out discussion of the Twitter argument with Sabine on this video, but you can find it on Twitter here - https://x.com/MLStreetTalk/status/1804825098207052061
https://x.com/Philip_Goff
Why? The Purpose of the Universe (Prof Philip Goff)
amzn.to/4cbYHqL
Galileo's Error: Foundations for a New Science of Consciousness
amzn.to/4eqA5Mi
Consciousness and Fundamental Reality
amzn.to/3VDd8wX
Is Consciousness Everywhere?
amzn.to/3xAiC3m
TOC
00:00:00 Introduction to Consciousness
00:02:52 Panpsychism Explained
00:04:43 Cosmopsychism and Panagentialism
00:08:44 Exploring Agency and Purpose
00:22:51 Physicalism vs. Panpsychism
00:28:03 Critique of Philosophical Views on Consciousness
00:30:51 Frank Jackson's Knowledge Argument
00:33:50 Galileo and the Foundations of Physical Science
00:36:43 Philosophical and Scientific Integration
00:43:57 MindChat and Constructive Disagreement
00:46:29 Book signing and Final Thoughts
We used a clip of Sam Harris from the Know Thyself podcast - youtube.com/watch?v=gqA-ZRpl1jQ
Jack and Mohammed explain their winning approach, which involves fine-tuning a language model on a large, specifically-generated dataset and then doing additional fine-tuning at test-time, a technique known in this context as "active inference". They use various strategies to represent the data for the language model and believe that with further improvements, the accuracy could reach above 50%. Michael talks about his work on generating new ARC-like tasks to help train the models.
They also debate whether their methods stay true to the "spirit" of Chollet's measure of intelligence. Despite some concerns, they agree that their solutions are promising and adaptable for other similar problems.
Note:
Jack's team is still the current official winner at 33% on the private set. Ryan's entry is not on the private leaderboard or eligible.
Chollet invented ARC in 2019 (not 2017 as stated)
"Ryan's entry is not a new state of the art. We don't know exactly how well it does since it was only evaluated on 100 tasks from the evaluation set and does 50% on those, reportedly. Meanwhile Jacks team i.e. MindsAI's solution does 54% on the entire eval set and it is seemingly possible to do 60-70% with an ensemble"
Jack Cole:
https://x.com/Jcole75Cole
https://lab42.global/community-interview-jack-cole/
Mohamed Osman:
Mohamed is looking to do a PhD in AI/ML, can you help him?
Email: mothman198@outlook.com
linkedin.com/in/mohamedosman1905
Michael Hodel:
arxiv.org/pdf/2404.07353v1
linkedin.com/in/michael-hodel
https://x.com/bayesilicon
github.com/michaelhodel
Getting 50% (SoTA) on ARC-AGI with GPT-4o - Ryan Greenblatt
redwoodresearch.substack.com/p/getting-50-sota-on-arc-agi-with-gpt
Neural networks for abstraction and reasoning: Towards broad generalization in machines [Mikel Bober-Irizar, Soumya Banerjee]
arxiv.org/pdf/2402.03507
Measure of intelligence:
arxiv.org/abs/1911.01547
I think the audio levelling might be a bit off on this for the intro especially, I fixed it on the audio podcast version - sorry if it's annoying.
Pod version: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/New-50-ARC-result-and-current-winners-interviewed-e2l1prl
TOC (autogenerated):
00:00:00 Introduction
00:03:00 Francois Chollet's Intelligence Concept
00:08:00 Human Collaboration
00:15:00 ARC Tasks and Symbolic AI
00:27:00 Evaluation Techniques
00:35:23 (Main Interview) Competitors and Approaches
00:40:00 Meta Learning Challenges
00:48:00 System 1 vs System 2
01:00:00 Inductive Priors and Symbols
01:18:00 Methodologies Comparison
01:25:00 Training Data Size Impact
01:35:00 Generalization Issues
01:47:00 Techniques for AI Applications
01:56:00 Model Efficiency and Scalability
02:10:00 Task Specificity and Generalization
02:13:00 Summary
Nick talks about his journey at Google Brain, working with AI legends like Geoff Hinton, and the amazing things his company, Cohere, is doing. From creating the must useful language models for businesses to making tools for developers, Nick shares a lot of interesting insights. He even talks about his band, Good Kid! Nick said that RAG is one of the best features of Cohere's new Command R* models. We are about to release a deep-dive on RAG with Patrick Lewis from Cohere, keep an eye out for that - he explains why their models are specifically optimised for RAG use cases.
Learn more about Cohere Command R models here:
cohere.com/command
github.com/cohere-ai/cohere-toolkit
Nick's band Good Kid:
goodkidofficial.com
Nick on Twitter:
https://x.com/nickfrosst
00:00:00 Intro
00:01:55 Backstory of Cohere
00:02:31 Hinton
00:02:54 Nick's band
00:03:11 How is Cohere differentiated?
00:03:44 Not an AGI company
00:06:00 Command+R
00:06:41 Standout feature: RAG
00:09:07 How is RAG changing the way we build apps?
00:09:44 Build day
00:10:25 Building robust applications
00:11:59 RAG evolution
00:14:45 Unsupervised RAG
00:16:30 Agents and divergence
00:18:27 Agency
00:22:19 Are LLMs general?
00:24:48 Benchmarks
00:27:07 Would Cohere verticalize?
00:27:43 RAG vs long context
00:29:20 Tech hasn't landed yet?
00:31:36 Are LLMs saturating?
00:35:50 Cohere's data acquisition pipeline
00:36:34 SOTA chasing vs fairness
00:37:21 Fall of the data scientist
00:40:37 Final callouts
Disclaimer: This is the first video from our Cohere partnership. We were not told what to say in the interview, and didn't edit anything out from the interview.
Please support us on Patreon to get access to the private Discord server, bi-weekly calls, early access and ad-free listening.
patreon.com/mlst
Aman Bhargava from Caltech and Cameron Witkowski from the University of Toronto to discuss their groundbreaking paper, “What’s the Magic Word? A Control Theory of LLM Prompting.” (the main theorem on self-attention controllability was developed in collaboration with Dr. Shi-Zhuo Looi from Caltech).
They frame LLM systems as discrete stochastic dynamical systems. This means they look at LLMs in a structured way, similar to how we analyze control systems in engineering. They explore the “reachable set” of outputs for an LLM. Essentially, this is the range of possible outputs the model can generate from a given starting point when influenced by different prompts. The research highlights that prompt engineering, or optimizing the input tokens, can significantly influence LLM outputs. They show that even short prompts can drastically alter the likelihood of specific outputs. Aman and Cameron’s work might be a boon for understanding and improving LLMs. They suggest that a deeper exploration of control theory concepts could lead to more reliable and capable language models.
We dropped an additional, more technical video on the research on our Twitter account here: https://x.com/MLStreetTalk/status/1795093759471890606
Pod version with no music/SFX:
podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Whats-the-Magic-Word--A-Control-Theory-of-LLM-Prompting-e2khs2t
Additional 20 minutes of unreleased footage on our Patreon here: patreon.com/posts/whats-magic-word-104922629
What's the Magic Word? A Control Theory of LLM Prompting (Aman Bhargava, Cameron Witkowski, Manav Shah, Matt Thomson)
arxiv.org/abs/2310.04444
LLM Control Theory Seminar (April 2024)
youtube.com/watch?v=9QtS9sVBFM0
Society for the pursuit of AGI (Cameron founded it)
agisociety.mydurable.com
Roger Federer demo
http://conway.languagegame.io/inference
Neural Cellular Automata, Active Inference, and the Mystery of Biological Computation (Aman)
aman-bhargava.com/ai/neuro/neuromorphic/2024/03/25/nca-do-active-inference.html
Aman and Cameron also want to thank Dr. Shi-Zhuo Looi and Prof. Matt Thomson from from Caltech for help and advice on their research. (https://thomsonlab.caltech.edu/ and https://pma.caltech.edu/people/looi-shi-zhuo)
https://x.com/ABhargava2000
https://x.com/witkowski_cam
TOC:
00:00:00 - Main Intro
00:06:25 - Bios
00:07:50 - Control Theory and Governors
00:09:37 - LLM Control Theory
00:17:17 - Federer Game
00:19:49 - Building LLM Controllers
00:20:56 - Priors in LLMs
00:28:44 - Manipulating LLMs
00:34:11 - Adversarial Examples and Robustification
00:36:54 - Model vs Software
00:39:12 - Experiments in the Paper
00:44:36 - Language as an Interstate Freeway
00:46:41 - Collective Intelligence
00:58:54 - Biomimetic Intelligence
01:03:37 - Society for the Pursuit of AGI
01:05:47 - ICLR Rejection
Throughout the conversation, Maria highlights her concern about the overreliance on AI in critical sectors such as healthcare, the justice system, and business. She stresses that while AI can serve as a tool, it should not replace human judgment and decision-making. Maria points out that AI systems often operate on past data, which may lead to outdated or incorrect decisions if not carefully managed.
The discussion also touches upon the concept of "adaptive resilience", which Maria describes in her book. She explains adaptive resilience as the capacity for individuals and enterprises to evolve and thrive amidst challenges by leveraging technology responsibly, without undermining human values and capabilities.
A significant portion of the conversation focussed on ethical considerations surrounding AI. Tim and Maria agree that there's a pressing need for strong governance and ethical frameworks to guide AI development and deployment. They discuss how AI, without proper ethical considerations, risks exacerbating issues like privacy invasion, misinformation, and unintended discrimination.
Maria is skeptical about claims of achieving Artificial General Intelligence (AGI) or a technological singularity where machines surpass human intelligence in all aspects. She argues that such scenarios neglect the complex, dynamic nature of human intelligence and consciousness, which cannot be fully replicated or replaced by machines.
Tim and Maria discuss the importance of keeping human agency and creativity at the forefront of technology development. Maria asserts that efforts to automate or standardize complex human actions and decisions are misguided and could lead to dehumanizing outcomes. They both advocate for using AI as an aid to enhance human capabilities rather than a substitute.
In closing, Maria encourages a balanced approach to AI adoption, urging stakeholders to prioritize human well-being, ethical standards, and societal benefit above mere technological advancement. The conversation ends with Maria pointing people to her book for more in-depth analysis and thoughts on the future interaction between humans and technology.
Buy Maria's book here: amzn.to/4avF6kq
linkedin.com/in/mariasantacaterina
POD: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/CAN-MACHINES-REPLACE-US--AI-vs-Humanity---Maria-Santacaterina-e2jaa4p
TOC
00:00:00 - Intro to Book
00:03:23 - What Life Is
00:10:10 - Agency
00:18:04 - Tech and Society
00:21:51 - System 1 and 2
00:22:59 - We Are Being Pigeonholed
00:30:22 - Agency vs Autonomy
00:36:37 - Explanations
00:40:24 - AI Reductionism
00:49:50 - How Are Humans Intelligent
01:00:22 - Semantics
01:01:53 - Emotive AI and Pavlovian Dogs
01:04:05 - Technology, Social Media and Organisation
01:18:34 - Systems Are Not That Automated
01:19:33 - Hiring
01:22:34 - Subjectivity in Orgs
01:32:28 - The AGI Delusion
01:45:37 - GPT-laziness Syndrome
01:54:58 - Diversity Preservation
01:58:24 - Ethics
02:11:43 - Moral Realism
02:16:17 - Utopia
02:18:02 - Reciprocity
02:20:52 - Tyranny of Categorisation
Interviewer: Dr. Tim Scarfe
Aman Bhargava, Cameron Witkowski, Manav Shah, Matt Thomson
arxiv.org/abs/2310.04444
Video out on Patreon today!
Active inference, developed by the legendary neuroscientist Prof. Karl Friston - is a unifying mathematical framework which frames living systems as agents which minimize surprise and free energy in order to resist entropy and persist over time. It unifies various perspectives from physics, biology, statistics, and psychology - and allows us to explore deep questions about agency, biology, causality, modelling, and consciousness.
Buy Active Inference: The Free Energy Principle in Mind, Brain, and Behavior
Thomas Parr, Giovanni Pezzulo, Karl Friston
amzn.to/4dj0iMj
Please support us on Patreon to get access to the private Discord server, bi-weekly calls, early access and ad-free listening.
patreon.com/mlst
Pod: podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/Dr--Thomas-Parr---Active-Inference-Book-e2j4sdn
TOC:
00:00:00 Intro
00:05:10 When Thomas met Friston
00:06:13 ChatGPT comparison
00:08:40 Do NNs learn a world model?
00:11:04 Book intro
00:13:22 High road low road of Active Inference
00:17:16 Resisting entropic forces
00:20:51 Agency vs free will
00:26:01 Are agents real? non-physical agents
00:35:54 Mind is flat / predictive brain
00:44:23 Volition
00:50:26 Externalism
00:51:57 Bridge with Enactivism
00:53:27 Bayesian Surprise
01:01:47 Variational inference
01:05:47 Why Bayesian?
01:12:04 Causality
01:17:35 Hand crafted models
01:26:45 Chapter 10 - bringing it together
01:28:58 Consciousness
01:33:10 Humans are incoherent
01:35:25 Experience writing a book
Interviewer: Dr. Tim Scarfe
Support MLST:
Please support us on Patreon. We are entirely funded from Patreon donations right now. Patreon supports get private discord access, biweekly calls, very early-access + exclusive content and lots more.
patreon.com/mlst
Donate: paypal.com/donate/?hosted_button_id=K2TYRVPBGXVNA
If you would like to sponsor us, so we can tell your story - reach out on mlstreettalk at gmail
Topics:
Externalized cognition and the role of society and culture in human intelligence
The potential for AI systems to develop agency and autonomy
The future of AGI as a complex mixture of various components
The concept of agency and its relationship to power
The importance of coherence in AI systems
The balance between coherence and variance in exploring potential upsides
The role of dynamic, competent, and incorruptible institutions in handling risks and developing technology
Concerns about AI widening the gap between the haves and have-nots
The concept of equal access to opportunity and maintaining dynamism in the system
Leahy's perspective on life as a process that "rides entropy"
The importance of distinguishing between epistemological, decision-theoretic, and aesthetic aspects of morality (inc ref to Hume's Guillotine)
The concept of continuous agency and the idea that the first AGI will be a messy admixture of various components
The potential for AI systems to become more physically embedded in the future
The challenges of aligning AI systems and the societal impacts of AI technologies like ChatGPT and Bing
The importance of humility in the face of complexity when considering the future of AI and its societal implications
TOC:
00:00:00 Intro
00:00:56 Connor's Philosophy
00:03:53 Office Skit
00:05:08 Connor on e/acc and Beff
00:07:28 Intro to Daniel's Philosophy
00:08:35 Connor on Entropy, Life, and Morality
00:19:10 Connor on London
00:20:21 Connor Office Interview
00:20:46 Friston Patreon Preview
00:21:48 Why Are We So Dumb?
00:23:52 The Voice of the People, the Voice of God / Populism
00:26:35 Mimetics
00:30:03 Governance
00:33:19 Agency
00:40:25 Daniel Interview - Externalised Cognition, Bing GPT, AGI
00:56:29 Beff + Connor Bonus Patreons Interview
Disclaimer: this video is not an endorsement of e/acc or AGI agential existential risk from us - the hosts of MLST consider both of these views to be quite extreme. We seek diverse views on the channel.
At Microsoft Research, Chris oversees a global portfolio of industrial research and development, with a strong focus on machine learning and the natural sciences.
Chris obtained a BA in Physics from Oxford, and a PhD in Theoretical Physics from the University of Edinburgh, with a thesis on quantum field theory.
Chris's contributions to the field of machine learning have been truly remarkable. He has authored (what is arguably) the original textbook in the field - 'Pattern Recognition and Machine Learning' (PRML) which has served as an essential reference for countless students and researchers around the world, and that was his second textbook after his highly acclaimed first textbook Neural Networks for Pattern Recognition.
Recently, Chris has co-authored a new book with his son, Hugh, titled 'Deep Learning: Foundations and Concepts.' This book aims to provide a comprehensive understanding of the key ideas and techniques underpinning the rapidly evolving field of deep learning. It covers both the foundational concepts and the latest advances, making it an invaluable resource for newcomers and experienced practitioners alike.
Buy Chris' textbook here:
amzn.to/3vvLcCh
More about Prof. Chris Bishop:
en.wikipedia.org/wiki/Christopher_Bishop
microsoft.com/en-us/research/people/cmbishop
Support MLST:
Please support us on Patreon. We are entirely funded from Patreon donations right now. Patreon supports get private discord access, biweekly calls, early-access + exclusive content and lots more.
patreon.com/mlst
Donate: paypal.com/donate/?hosted_button_id=K2TYRVPBGXVNA
If you would like to sponsor us, so we can tell your story - reach out on mlstreettalk at gmail
TOC:
00:00:00 - Intro to Chris
00:06:54 - Changing Landscape of AI
00:08:16 - Symbolism
00:09:32 - PRML
00:11:02 - Bayesian Approach
00:14:49 - Are NNs One Model or Many, Special vs General
00:20:04 - Can Language Models Be Creative
00:22:35 - Sparks of AGI
00:25:52 - Creativity Gap in LLMs
00:35:40 - New Deep Learning Book
00:39:01 - Favourite Chapters
00:44:11 - Probability Theory
00:45:42 - AI4Science
00:48:31 - Inductive Priors
00:58:52 - Drug Discovery
01:05:19 - Foundational Bias Models
01:07:46 - How Fundamental Is Our Physics Knowledge?
01:12:05 - Transformers
01:12:59 - Why Does Deep Learning Work?
01:16:59 - Inscrutability of NNs
01:18:01 - Example of Simulator
01:21:09 - Control
He trained as a chemist at the University of Oxford, and as a physicist at the University of Bristol. He worked previously at Nature for over 20 years, first as an editor for physical sciences and then as a consultant editor. His writings on science for the popular press have covered topical issues ranging from cosmology to the future of molecular biology.
Philip is the author of many popular books on science, including H2O: A Biography of Water, Bright Earth: The Invention of Colour, The Music Instinct and Curiosity: How Science Became Interested in Everything. His book Critical Mass won the 2005 Aventis Prize for Science Books, while Serving the Reich was shortlisted for the Royal Society Winton Science Book Prize in 2014.
This is one of Tim's personal favourite MLST shows, so we have designated it a special edition. Enjoy!
Buy Philip's book "How Life Works" here: amzn.to/3vSmNqp
TOC:
00:00:00 Outside interview
00:09:17 Nativism / capacities
00:11:50 Generative AI
00:17:59 Inscrutability and agency
00:22:06 Agency on creativity
00:26:38 Could we make an agential GPT-4?
00:31:06 If it agential if you tell the agents what to do
00:35:40 What is agency?
00:44:29 Are agents real
00:48:32 Causality and agency
00:54:40 Ghost in the machine / intelligibility
01:00:11 Multi scale / organisation view
01:04:05 Collective intelligence
01:09:00 Canalisation
01:13:36 Intelligence is specialised
01:16:29 No free lunch
01:18:19 Super intelligence
01:22:05 Mind is flat / confabulated goals
01:25:07 Is planning/goals explicit?
01:30:17 Sentience in LLMs
01:34:56 Are LLMs simulators?
01:40:04 Could LLMs feel?
01:49:33 Digital physics view
01:51:46 Agential vs nonagential AI
01:54:02 Bostrom thinks goals and intelligence are separate
02:05:41 Simulation sharing
AI Transcript: docs.google.com/document/d/164FMAswgS9jAqxAaB5fkUtrJKMdq8rW7b94XgibvQ7Q/edit?usp=sharing
Support MLST:
Please support us on Patreon. We are entirely funded from Patreon donations right now. Patreon supports get private discord access, biweekly calls, early-access + exclusive content and lots more.
patreon.com/mlst
Donate: paypal.com/donate/?hosted_button_id=K2TYRVPBGXVNA
If you would like to sponsor us, so we can tell your story - reach out on mlstreettalk at gmail