Anthropic just proved AI isnt getting better Mo Bitar 2026-09-19 | Anthropic just proved AI isnt getting better
AI isn’t taking jobs. It’s taking something worse. Mo Bitar 2026-04-27 | You losing your job is the best thing to ever happen.This was a member-only video that my members urged me to make public :) Please forgive the excessive profanity.Precision/utility relationship: https://x.com/atmoio/status/2048139751933309154https://x.com/atmoio/status/2048608025464705084Employees at Meta are tracked based on how much tokens they consume: fortune.com/2026/04/09/meta-killed-employee-ai-token-dashboarden.wikipedia.org/wiki/Bitter_lesson?wprov=sfti1businessinsider.com/jensen-huang-500k-engineers-250k-ai-tokens-nvidia-compute-2026-3Adjacent OpenCode tweet, I can't find the original one: https://x.com/thdxr/status/2022574719694758147atmoio.substack.comintro.co/MoBitarEquipment I use:Camera: amzn.to/3QyjE96Boom mic: amzn.to/4d0GbnPAudio interface: amzn.to/4d0GbnPHandheld mic: amzn.to/4crpRvcLens: amzn.to/41Qfqw7Key light: amzn.to/3QGohy5Light dome: amzn.to/4tnCGxzSide lights: amzn.to/3QhATf4Middle back light: amzn.to/4mMtZKWPainting lights: amzn.to/4vMROWZ
Microsoft accidentally told the truth about AI Mo Bitar 2026-04-23 | Rogue researchers are telling the truth about AIDepth vs breadth: https://x.com/atmoio/status/2041557482217120182Make ze slop: https://x.com/atmoio/status/2046421781326500029atmoio.substack.comintro.co/MoBitarEquipment I use:Camera: amzn.to/3QyjE96Boom mic: amzn.to/4d0GbnPAudio interface: amzn.to/4d0GbnPHandheld mic: amzn.to/4crpRvcLens: amzn.to/41Qfqw7Key light: amzn.to/3QGohy5Light dome: amzn.to/4tnCGxzSide lights: amzn.to/3QhATf4Middle back light: amzn.to/4mMtZKWPainting lights: amzn.to/4vMROWZSources:https://x.com/venturetwins/status/2046071130797285668https://www.wheresyoured.at/news-microsoft-to-shift-github-copilot-users-to-token-based-billing-reduce-rate-limits-2/cnn.com/2026/04/08/tech/ai-software-developer-jobscitadelsecurities.com/news-and-insights/2026-global-intelligence-crisisyoutube.com/watch?v=TYuSEeuUhPotheregister.com/2026/03/31/anthropic_claude_code_limitsgithub.com/anthropics/claude-code/issues/41788spectrumailab.com/blog/claude-code-vs-cursorarxiv.org/abs/2604.15597arxiv.org/abs/2506.08872futurism.com/neoscope/brain-eeg-chatgpt--LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
I know how the AI bubble ends now Mo Bitar 2026-04-17 | We've reached the scam singularity.If you'd like to support my work, you can become a member to view member-only videos, where I post videos that go deeper on some of the topics we discuss here.youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/join Thank you!!https://x.com/atmoioatmoio.substack.comintro.co/MoBitarEquipment I use:Camera: amzn.to/3QyjE96Mic: amzn.to/4crpRvcLens: amzn.to/41Qfqw7Key light: amzn.to/3QGohy5Light dome: amzn.to/4tnCGxzMiddle back light: amzn.to/4mMtZKWPainting lights: amzn.to/4vMROWZSources: ir.allbirds.com/news-releases/news-release-details/allbirds-inc-executes-50m-convertible-financing-facilityfinance.yahoo.com/news/allbirds-stock-on-track-for-400-gain-this-week-as-shoemaker-pivots-to-ai-133721433.htmlsec.gov/newsroom/press-releases/2021-121---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
Harvard just discovered what AI actually is Mo Bitar 2026-04-14 | You're absolutely right, that IS a great reason to fire everyone!If you'd like to support my work, you can become a member to view member-only videos, where I post videos that go deeper on some of the topics we discuss here.youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/join Thank you!!https://x.com/atmoio/status/2021973014775312496https://x.com/atmoioatmoio.substack.comEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96Sources:hbr.org/2026/03/researchers-asked-llms-for-strategic-advice-they-got-trendslop-in-returnfinance.yahoo.com/news/gaming-ceo-asked-chatgpt-avoid-192340700.htmlinc.com/ben-sherry/is-your-business-idea-actually-good-this-claude-hack-provides-the-cold-hard-truth/91328813en.wikipedia.org/wiki/Krafton404media.co/ceo-ignores-lawyers-asks-chatgpt-how-to-void-250-million-contract-loses-terribly-in-court---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
Claude Mythos is Delusional Mo Bitar 2026-04-08 | Meet the new AI model with its own therapist.Become a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoioatmoio.substack.comEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96Sources:anthropic.com/glasswingtechcrunch.com/2019/02/17/openai-text-generator-dangerous---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
MIT Just Proved AI Can’t Do Your Job Mo Bitar 2026-04-06 | You're absolutely right, I shouldn't have put you in jail!Become a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoio/status/2040797533568434350atmoio.substack.comEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96Sources:semafor.com/article/11/26/2025/deloitte-faces-new-scrutiny-over-ai-generated-mistakescbc.ca/news/canada/newfoundland-labrador/nl-deloitte-citations-9.6990216fastcompany.com/91417492/deloitte-ai-report-australian-governmentfortune.com/2025/10/07/deloitte-ai-australia-government-report-hallucinations-technology-290000-refundtheguardian.com/australia-news/2025/oct/06/deloitte-to-pay-money-back-to-albanese-government-after-using-ai-in-440000-reportcnbc.com/2026/02/06/google-microsoft-meta-amazon-ai-cash.htmlarxiv.org/pdf/2604.01363tech.yahoo.com/ai/articles/mit-created-duplicate-ai-workers-185644013.html?guccounter=1microsoft.com/en-us/microsoft-copilot/for-individuals/termsofuse---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
The AI crisis no one is talking about Mo Bitar 2026-04-03 | Asking ChatGPT about pi was the worst mistake he ever made.Become a member on YouTube:youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinBook time with me:intro.co/MoBitarThis video was inspired by this tweet:https://x.com/atmoio/status/2039426909440725158Subscribe to Substack, where I share a written rendition of these videos:atmoio.substack.comSourcesarxiv.org/pdf/2602.19141techcrunch.com/2025/06/15/spiraling-with-chatgptnytimes.com/2025/08/08/technology/ai-chatbots-delusions-chatgpt.htmlcbc.ca/news/canada/ai-psychosis-canada-1.7631925torontolife.com/deep-dives/man-vs-machine-chatgpt-delusion-now-hes-suing-openaifuturism.com/chatgpt-chabot-severe-delusionsnytimes.com/2025/06/13/technology/chatgpt-ai-chatbots-conspiracies.htmlthehumanlineproject.org---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
Normal people are starting to go crazy Mo Bitar 2026-03-27 | You're absolutely right, you ARE a prophet!Become a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoioatmoio.substack.comEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96My sincere thanks to Fish for sponsoring this video. Fish is an open-source, open-weights synthetic voice tool that is "the most expressive, emotionally controllable real-time voice model."Use this link and enter discount code Mo20 to get started: https://fish.audio/?fpr=mo62(This is a non-affiliate link, and I make no commission. It's used only to track that you came from this video 🙂)---Sources:futurism.com/future-society/openai-data-chatgpt-mental-health-criseswired.com/story/chatgpt-psychosis-and-self-harm-updatehttps://x.com/EricRWeinstein/status/2036156032003744198---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm wholly convinced is possible.
The internet is dying Mo Bitar 2026-03-24 | The barrier to destroying the internet is now zero. Thanks OpenClaw.Become a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoio/status/2034693261231566858Equipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96---https://x.com/atmoioatmoio.substack.comintro.co/MoBitarSources:https://x.com/nikitabier/status/2021632774013432061https://x.com/nikitabier/status/2024729689156440326https://x.com/nikitabier/status/2017134769113542752dexerto.com/entertainment/x-is-testing-a-dislike-button-again-and-its-coming-soon-3336926huggingface.co/blog/rlhfbbc.com/news/uk-24992393theguardian.com/technology/2025/aug/11/cat-soap-operas-and-babies-trapped-in-space-the-ai-slop-taking-over-youtubepressgazette.co.uk/media-audience-and-business-data/google-traffic-down-2025-trends-report-2026blogs.nvidia.com/blog/gtc-2026-news---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm wholly convinced is possible.
ChatGPT Adult Mode is hilariously bad Mo Bitar 2026-03-21 | As an AI language model, I'm not able to—just kidding take off your pants.https://x.com/atmoioatmoio.substack.comintro.co/MoBitarSources:wsj.com/tech/ai/openai-adult-mode-chatgpt-f9e5fc1atechcrunch.com/2025/10/27/openai-says-over-a-million-people-talk-to-chatgpt-about-suicide-weeklysfstandard.com/2025/11/04/sam-altman-steve-kerr-san-francisco-talknypost.com/2025/11/08/us-news/openai-ceo-sam-altman-served-subpoena-onstage-during-san-francisco-talkopenai.com/index/expert-council-on-well-being-and-aihttps://x.com/sama/status/1978129344598827128arstechnica.com/tech-policy/2026/03/chatgpt-may-soon-become-sexy-suicide-coach-openai-advisor-reportedly-warnednytimes.com/2025/08/26/technology/chatgpt-openai-suicide.html?partner=slack&smid=sl-shareen.wikipedia.org/wiki/Deaths_linked_to_chatbotsfuturism.com/artificial-intelligence/chatgpt-suicide-openai-gpt4oainvest.com/news/openai-14-billion-2026-loss-burn-priced-2603engadget.com/ai/openais-adult-mode-reportedly-wont-generate-pornographic-audio-images-or-video-150744035.htmlpolitico.com/news/2026/02/26/anthropic-rejects-pentagons-ai-demands-00802554---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images, synthetic voices (David Attenborough and Scarlett Johansson), and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm wholly convinced is possible.
OpenAI is finding ChatGPT useless Mo Bitar 2026-03-19 | The company that promised AI would do the job of 10 people, can't even do the job of ONE company.Become a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoioatmoio.substack.comintro.co/MoBitarEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96Articles referenced in this video:techcrunch.com/2025/07/31/enterprises-prefer-anthropics-ai-models-over-anyone-elses-including-openaisfuturism.com/artificial-intelligence/openai-cutting-projectswired.com/story/openai-codex-race-claude-codenytimes.com/2026/02/12/opinion/artificial-intelligence-anthropic-amodei.htmltechnologyreview.com/2025/10/27/1126673/openai-new-atlas-browsernytimes.com/2025/10/03/technology/sora-openai-video-disinformation.htmlwsj.com/tech/ai/openai-chatgpt-side-projects-16b3a825wsj.com/tech/ai/anthropic-claude-code-ai-7a46460emetr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-studyhttps://www.index.dev/blog/developer-productivity-statistics-with-ai-tools---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm wholly convinced is possible.
AI is making CEOs delusional Mo Bitar 2026-03-16 | Your CEO tries AI once over the weekend and by Monday your company is "AI-first" smhBecome a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoioatmoio.substack.comintro.co/MoBitarEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm wholly convinced is possible.
Amazon is regretting AI Mo Bitar 2026-03-13 | You're absolutely right, I shouldn't have deleted production!Become a member to view exclusive member-only videos, and support my work :) youtube.com/channel/UCAwT-7B6StW_xzXPku6v7MQ/joinhttps://x.com/atmoioatmoio.substack.comintro.co/MoBitarEquipment I use:Mic: amzn.to/4crpRvcCamera: amzn.to/3QyjE96---LLMs were used on a very short leash to assist in the production of this video to: generate b-roll images and deep research. Narrative beginning, middle, and end, as well as color, analysis, conclusion, and most importantly the comic relief, are always written by me. Occasionally, I might keep elements from the LLM that I thought were useful to advance the narrative structure. It's important to note that I am not anti-AI, but rather anti being-stupid-with-AI (and stupidity abounds, which is what creates the opportunity to make these videos). One would be quite mistaken to treat this technology as completely useless. My consistent take on this channel has been that AI is supremely useful, but is not a tool that can automate or replace humans. It must be kept on a very tight leash. This limitation is not temporary, but rather inherit to LLMs. AGI would solve these problems, but is not something I'm convinced is possible.
I was a 10x engineer. Now Im useless. Mo Bitar 2026-03-07 | AI is frying my brain. Is anyone else feeling this?https://x.com/@atmoiohttps://shape.work
Cloudflare just broke all the rules Mo Bitar 2026-03-04 | A single engineer slopforked the internet's most valuable framework from Vercel. The future of vibecoding just got real.https://x.com/@atmoio
Anthropic nearly started a war. Here’s how to make sense of their strategy. Mo Bitar 2026-03-03 | Anthropic’s psychosis around Claude starts to make sense when you consider them in this perspective.https://x.com/atmoio/status/2028829192465379483
Companies are lying about AI Mo Bitar 2026-02-27 | Jack Dorsey's company Block/Square laid off 4,000 and Jack blamed it on AI. But there's something important you need to understand about AI and your job safety.https://x.com/jack/status/2027129697092731343https://x.com/atmoio/status/202306555729825010700:00 Jack is lying04:17 Please watch this part before you leave.I'm available 1:1 if I can help you with anything, including entrepreneurship, strategy, software development, startup creation, advice, inspiration, feedback, career navigation, etc: intro.co/MoBitar
Anthropic is seriously tripping Mo Bitar 2026-02-26 | San Francisco software engineers at Anthropic are having a hard time letting go.https://x.com/AnthropicAI/status/2026765820098130111https://x.com/atmoioI'm available for a call if I can help you with anything 1:1, including entrepreneurship, strategy, software development, startup creation, advice, inspiration, feedback, career navigation, etc: intro.co/MoBitar
Cursor just solved vibecoding. But its very scary. Mo Bitar 2026-02-25 | Shipping unreviewed code is immoral and dangerous.https://x.com/cursor_ai/status/2026369873321013568https://x.com/atmoioI'm available for a call if I can help you with anything 1:1, including entrepreneurship, strategy, software development, startup creation, advice, inspiration, feedback, career navigation, etc: intro.co/MoBitar
6-figure software engineers are now driving Ubers Mo Bitar 2026-02-24 | There's one thing you can do that an AI can never.https://x.com/Citrini7/status/2025653614430023864https://x.com/atmoio/status/2026030907467940088I'm available for a call if I can help you with anything 1:1, including entrepreneurship, strategy, software development, startup creation, advice, inspiration, feedback, career navigation, etc: intro.co/MoBitar
They want to drop the AI nuke on YOU Mo Bitar 2026-02-21 | No datacenters, no AI.https://x.com/BenDziobek/status/2024298250203750567https://x.com/Charlie4Change/status/2024346491314110507https://x.com/WillManidis/status/2024938679899168855
Will AI keep getting better? Mo Bitar 2026-02-20 | It ain't easy being an AI skeptic these days.- Stripe: https://x.com/stripe/status/2024574740417970462- LLMs: https://x.com/atmoio/status/2023065557298250107- Genie: https://x.com/atmoio/status/2024130241707331993- Sundar: https://x.com/sundarpichai/status/2024516418855981298
AI is not ready. Mo Bitar 2026-02-16 | AI is not ready, and it's being shoved down your throat by largely 1-2 people. https://x.com/notch/status/2023006241090613454https://x.com/atmoio/status/2022395850744565876
ChatGPT Pro is lapping Einstein Mo Bitar 2026-02-14 | OpenAI said "GPT‑5.2 derives a new result in theoretical physics," but this is highly misleading and sensationalized. I break down what actually happened, and what it means for researchers, engineers, artists, and you, going forward.- https://x.com/atmoio/status/2022410531928502481- openai.com/index/new-result-theoretical-physics
AI psychosis is spreading Mo Bitar 2026-02-11 | An article with 80M views on X is giving everyone LLM psychosis.https://x.com/atmoio/status/2021589614981460424https://x.com/mattshumer_/status/2021256989876109403
The truth behind AI “solving” the Erdos problems Mo Bitar 2026-02-10 | Did LLMs really make new discoveries in mathematics?https://x.com/atmoio/status/2020530273285197912newscientist.com/article/2511954-amateur-mathematicians-solve-long-standing-maths-problems-with-ai
10,000 lines of bug-free code a day!! #coding #vibecoding #software #codex Mo Bitar 2026-02-08 | Paul Graham met another founder. This one is writing 10k lines of bug-free code, and it's definitely not Garry Tan!
OpenAI says SLOP is here to STAY. #openai #coding #software #codex #vibecoding Mo Bitar 2026-02-07 | This is the most honest thing I've ever heard from an executive at an AI company.
Opus 4.6 is a slop machine Mo Bitar 2026-02-06 | These models are helping Garry Tan write 10k lines of bug-free code per day!https://x.com/atmoio/status/2019605765422809383https://x.com/gdb/status/2019566641491963946https://x.com/iannuttall/status/2019715935008174471https://x.com/ThePrimeagen/status/2019608375529787554
I don’t trust Anthropic Mo Bitar 2026-02-05 | There's something off about Anthropic.- forum.effectivealtruism.org/posts/6XbtL93kSFJwX45X2/unless-its-governance-changes-anthropic-is-untrustworthy- anthropic.com/constitution- https://x.com/sama/status/2019139174339928189- youtube.com/watch?v=lbITsEcSb6800:00 Intro04:15 Yann LeCun and Open Source07:03 Banning OpenAI and XAI07:46 "Unless its governance changes, Anthropic is untrustworthy"16:42 OpenCode, Clawdbot, and developer sentiment18:26 Claude's Constitution and Liability Control21:09 Dario's Essays22:28 Propaganda, narrative control, and you.
You can’t learn to code by vibecoding Mo Bitar 2026-02-04 | "Unpopular opinion: if you spend enough time vibe coding, you'll eventually learn to code."No. No no no no.If you want to learn to code (and you should), learn to code.Tweet: https://x.com/atmoio/status/2019069195888087248There is no skill in AI coding: atmoio.substack.com/p/there-is-no-skill-in-ai-coding
Vibecoders will never be software developers Mo Bitar 2026-02-03 | Many devs are afraid of losing their job to AI. But you can make yourself irreplaceable by following these 7 rules of what makes a great software engineer, and what differentiates an engineer from a vibecoder.Post: https://x.com/atmoio/status/2018686267026735388Timestamps:00:00 Intro01:18 An engineer can correct AI03:48 An engineer can instill confidence in management06:38 An engineer gets better and more valuable over time09:47 An engineer can evolve a product and organization13:19 An engineer can take responsibility for their work15:54 An engineer can lead other engineers20:28 An engineer can say no23:08 Summary
I hope you hold on to your humanity Mo Bitar 2026-02-01 | If you have been feeling confused or anxious about AI and what it all means, I hope this video helps you make sense of it all.I hope you come to understand that AI is a science-fiction term, and always will be."X’s narrative-search algorithm has over-indexed for many days now on the Clawdbot/Moltbot/OpenClaw phenomenon of agents seemingly coming to life. But there’s no need to wait for the Black Mirror episode to make sense of it all. What you see when you stare directly into the eyes of what’s happening is: nothing."Full article: atmoio.substack.com/p/i-hope-you-hold-on-to-your-humanityX post: https://x.com/atmoio/status/2017828766022574364?s=20
Theres no skill in AI coding Mo Bitar 2026-01-28 | I've begun to suspect something you have probably begun to suspect as well: there's no real skill in AI coding. The models are always the one doing the heavy lifting. You can't change them, no matter how clever you think you're being. The models are simply junior developers.There's no skill in AI coding article on X: https://x.com/atmoio/status/2016510083107066301On Substack: atmoio.substack.com/p/there-is-no-skill-in-ai-coding
AI coding will be a minimum wage job Mo Bitar 2026-01-26 | Be very careful taking advice like this from corporate CEOs: "Google DeepMind's CEO advises undergraduates that getting unbelievably proficient with AI tools is now more valuable than traditional internships for leapfrogging into a profession."So should you still learn to code? And should you get a real internship? Yes and yes.0:00 Should you get an internship?5:35 Should you learn to code?
After two years of vibecoding, I’m back to writing by hand #coding #vibecoding #software #claude Mo Bitar 2026-01-23 | You may not like it, but this is the truth of the matter.
After two years of vibecoding, I’m back to writing by hand Mo Bitar 2026-01-23 | You may not like it, but this is the truth of the matter.
The AI developer bubble is going to pop soon Mo Bitar 2026-01-22 | And when it does, people that can code will be the Chosen.
The AI developer bubble is going to pop soon - Shape Devlog - Day 5 Mo Bitar 2026-01-22 | Today:- Trying Google's Antigravity, which has a WAY better autocomplete model than Cursor- Refactoring smelly services into usecases- Why the AI developer bubble is going to pop soon, and why people that can code will be the Chosen.Timestamps0:00 - Service refactor10:44 - Antigravity is WAY better than Cursor15:00 - The AI developer bubble popping is going to be catastrophic
AI can’t mess up your code if there’s no code in the first place - Shape Devlog - Jan 21 Mo Bitar 2026-01-21 | I figured it out. I'm going to put my client, server, and engine in a single file. And put everything in a single table. Then neither me or AI will ever be confused again.Why hasn't anyone thought about this before?
AI tricks you into accepting bad code - Shape Devlog - Day 3 Mo Bitar 2026-01-20 | How much time did I really save having AI write my code, if I now have to go in and spend 14 hours a day correcting it all to be borderline presentable?And Cursor autocomplete is driving me nuts.I'm done man.
Services are a codesmell - Shape Devlog - Day 2 Mo Bitar 2026-01-17 | I got rid of the core user-service.ts and workspace-manager.ts in favor of modular usecases and tiny stores.Everything just clicked, the tests passed in their first run, and getting down and dirty with typescript library refactors makes me feel like I'm in 2022 again 🥹
Throwing away 10 hours of Codex work - Shape Devlog - Day 1 Mo Bitar 2026-01-16 | Experimenting with a new format. Enjoy!Today I worked on multiple user support for shape.work with Codex xhigh, high, and medium, and also Opus 4.5.
