Skip to content
Artwork for AI Explained
NewsTechnologyEducation

AI Explained

AI Explained

Covering the biggest news of the century - the arrival of smarter-than-human AI. The author of Simple Bench, exposing the remaining human-LLM reasoning gap.

Solo Developer of LM Council: https://lmcouncil.ai - Code - INSIDER15

Join me at AI Insiders, with exclusive videos and a 1000+ network of AI enthusiasts and professionals: https://www.patreon.com/AIExplained

Links:
AI Insiders: Code - INSIDER15: https://www.patreon.com/AIExplained
My AI Council App: https://lmcouncil.ai
Simple Bench: https://simple-bench.com/
Podcast (New!): https://aiexplainedopodcast.buzzsprout.com/
Newsletter: https://signaltonoise.beehiiv.com/
X: https://twitter.com/AIExplainedYT

Play
  • 145 episodes
  • daily
  • Avg 24 min
  • English
Counted on this page — what you have heard stays on this device, so it is not something the list can be paged by.
  • Yesterday · 29 min

    Gemini 2.5 Pro - It’s a Darn Smart Chatbot … (New Simple High Score)

    Gemini gets a new record on Simple Bench, and several other benchmarks. I’ll go deep to explore its nuances, including how it deceptively reverse engineers answers, does better on certain coding benchmarks than others, may have a universal ‘conceptual language’ … https://weave-docs.wandb.ai/?utm_source=sponsorship&utm_medium=simple_bench&utm_campaign=ai_explained … and more. Plus practical tips, a note on security and Kling vs Veo 2 guest appearance. AI Insiders ($9!): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 00:36 - Fiction Bench 02:41 - Practicality - YouTube urls + Security - cut-off date 03:42 - Coding 06:22 - WeirdML Bench 07:01 - Simple Bench Record High 11:23 - Reverse Engineering! 13:22 - Anthropic Paper 17:49 - 3 Caveats Gemini 2.5 Updated: https://deepmind.google/technologies/gemini/ Fiction Live Bench: https://fiction.live/stories/Fiction-liveBench-Feb-19-2025/oQdzQvKHw8JyXbN87 https://simple-bench.com/ WeirdML: https://htihle.github.io/weirdml.html https://x.com/htihle/status/1905014058228625542 Anthropic Thoughts: https://www.anthropic.com/research/tracing-thoughts-language-model https://transformer-circuits.pub/2025/attribution-graphs/biology.html#dives-cot https://aistudio.google.com/prompts/new_chat Search Study: https://www.cjr.org/tow_center/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php Live bench: https://livebench.ai/#/ Paper: https://arxiv.org/pdf/2406.19314 LiveCode Bench: https://livecodebench.github.io/ SWE-Verified: https://arxiv.org/pdf/2310.06770 Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 22 min

    New Google Model Ranked ‘No. 1 LLM’, But There’s a Problem

    A new and mysterious Gemini model appears at the top of the leaderboard, but is that the full story? I dig behind the headline to show you some anti-climactic results, give some context with leaks in the last 48 hours of diminishing returns to scaling, and add the response of Altman, OpenAI and co. The future is about to look a lot stranger. 80,000 hours Podcast + Channel: https://open.spotify.com/show/2WzJwXWBDnn4iZ7odKwDib https://www.youtube.com/@eightythousandhours/videos You can now gift memberships to AI Insiders (my Patreon w/ exclusive vids, network): https://www.patreon.com/AIExplained/gift Chapters: 00:00 - Introduction 01:25 - LM Leaderboard 02:35 - Benchmarks and Leaks 05:31 - Low EQ 07:37 - Other labs have issues too though 10:31 - OpenAI claim and counter-claim 14:13 - Other news ‘There is no wall’: https://x.com/sama/status/1856941766915641580 simple-bench.com https://x.com/vedantmisra/status/1857148554105544708 Gemini Ranking: https://lmarena.ai/?leaderboard API not yet up: https://x.com/OfficialLoganK/status/1857106844805681153 ‘Just Die Chat’: https://x.com/koltregaskes/status/1856754648146653428 https://gemini.google.com/share/6d141b742a13 Google CEO tweet: https://x.com/sundarpichai/status/1857114106928718329 Sutskever Quote: https://www.reuters.com/technology/artificial-intelligence/openai-rivals-seek-new-path-smarter-ai-current-methods-hit-limitations-2024-11-11/ Another OpenAI Staffer Leaves: https://x.com/RichardMCNgo/status/1856843040427839804 Bloomberg Report: https://www.bloomberg.com/news/articles/2024-11-13/openai-google-and-anthropic-are-struggling-to-build-more-advanced-ai?s=09 Noam Brown on what OpenAI Researchers Believe: https://x.com/polynoamial/status/1855037689533178289 Clive Chan: https://x.com/itsclivetime/status/1855704120495329667 Chollet Responds to Altman: https://x.com/fchollet/status/1857060079586975852 https://x.com/sama/status/1856940152460869718 Altman Emails: https://x.com/TechEmails/status/1857285960997712356 Change of Heart: https://sd11.senate.ca.gov/news/senator-wiener-responds-openai-opposition-sb-1047 Amodei on ‘Empirical Regularities’: https://lexfridman.com/dario-amodei-transcript/ Verge Report: https://www.theverge.com/2024/10/25/24279600/google-next-gemini-ai-model-openai-december OpenAI Agents in January: https://www.bloomberg.com/news/articles/2024-11-13/openai-nears-launch-of-ai-agents-to-automate-tasks-for-users?srnd=phx-ai The 8 Most Controversial Terms in AI: https://imp.i384100.net/m57g3M Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ I use Descript to edit my videos: https://get.descript.com/ldgxfuj2bhnb Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 32 min

    OpenAI Backtracks, Gunning for Superintelligence: Altman Brings His AGI Timeline Closer - '25 to '29

    Sam Altman unexpectedly brings his timelines to AGI forward, while OpenAI backtrack on superintelligence. None of these changes were heralded, but they are significant. Plus the new year brings new assessments of the true capability of models to automate 'large swathes of the economy'. I'll give my prediction on that front for 2025, announcement a new Simple Bench competition, and showcase Kling 1.6 vs Veo 2 vs Sora, and much more. wandb.me/simple-bench (Colab): https://colab.research.google.com/drive/1AVijcPnEkl8Gy_754XbRdG5m7Q5-9slg?usp=sharing AI Insiders: https://www.patreon.com/AIExplained TheAgentCompany Paper: https://arxiv.org/pdf/2412.14161v1 Sam Altman Major Interview: https://www.bloomberg.com/features/2025-sam-altman-interview/?srnd=phx-ai OpenAI Agent Coming Jan 2025: https://www.theinformation.com/articles/why-openai-is-taking-so-long-to-launch-agents?rc=sy0ihq Altman Singularity: https://x.com/sama/status/1875603249472139576 Altman Original Timeline: https://www.youtube.com/watch?v=7dCPytNTnjk&t=621s https://www.ft.com/content/34a7a082-e685-4e02-bca7-61ff89d99ed2 OpenAI Original Emails: https://www.lesswrong.com/posts/5jjk4CDnj9tA7ugxr/openai-email-archives-from-musk-v-altman-and-openai-blog DeepMind Sky News 2014 Article: https://news.sky.com/story/google-buys-uk-intelligence-firm-deepmind-10419783 Altman Blog Reflections: https://blog.samaltman.com/reflections OpenAI Changes Who Gets AGI: https://openai.com/index/why-our-structure-must-evolve-to-advance-our-mission/?s=09 OpenAI 5 Levels: https://www.bloomberg.com/news/articles/2024-07-11/openai-sets-levels-to-track-progress-toward-superintelligent-ai Altman 2015: https://blog.samaltman.com/machine-intelligence-part-1 OpenAI React to Anthropic: https://www.theinformation.com/articles/how-anthropic-got-inside-openais-head?rc=sy0ihq Microsoft $100B Definition: https://www.theinformation.com/articles/microsoft-and-openai-wrangle-over-terms-of-their-blockbuster-partnership?rc=sy0ihq Epoch Scramble for Task Benchmark: https://x.com/tamaybes/status/1876692639363612919 GPQA Progress: https://epoch.ai/data/ai-benchmarking-dashboard Task Length Crucial for ARC-AGI: https://anokas.substack.com/p/llms-struggle-with-perception-not-reasoning-arcagi RL Environment Tweet: https://x.com/vedantmisra/status/1876327518157807990 Jason Wei Talk: https://www.youtube.com/watch?v=yhpjpNXJDco Miles Brundage Tweet: https://x.com/Miles_Brundage/status/1872676399762612457 Jan Leike Tweet: https://x.com/janleike/status/1872909496777134127 O1 Pro Losing Money: https://x.com/sama/status/1876104315296968813 Kling 1.6: https://klingai.com/text-to-video/106634842 Chapters: 00:00 - Introduction 01:03 - Altman Timeline Moves Forward 04:33 - Superintelligence? 06:55 - AGI was not the only pitch 09:26 - AgentCompany and OpenAI New Agent 17:24 - SimpleBench Competition 23:03 - Kling 1.6 vs Veo 2 vs Sora Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 27 min

    8 Ways ChatGPT 4 [Is] Better Than ChatGPT

    Hear me now, quote me later: these are the 8 ways in which ChatGPT will be improved in ChatGPT 4. From critical reasoning to coding, comprehension to calculation, I will give the peer-reviewed evidence of what is to come. By examining publicly accessible benchmarks, comparable large language models and the latest research papers, we can discern the ways in which GPT4 (integrated into Bing or otherwise) will beat ChatGPT. I'll show you how unreleased models already beat current ChatGPT and all of this will actually give a clearer insight into what even GPT5 and rival models from Google might well be able to achieve. https://ai.googleblog.com/2022/04/pathways-language-model-palm-scaling-to.html https://cloud.google.com/blog/topics/tpus/google-showcases-cloud-tpu-v4-pods-for-large-model-training https://arxiv.org/pdf/2204.02311.pdf https://arxiv.org/pdf/1905.00537.pdf https://arxiv.org/pdf/2201.11903.pdf https://github.com/google/BIG-bench/ https://github.com/google/BIG-bench/blob/main/bigbench/benchmark_tasks/README.md https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/gre_reading_comprehension https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/logical_args https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/evaluating_information_essentiality https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/physics http://web.mit.edu/~yczeng/Public/WORKBOOK%201%20FULL.pdf https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/sufficient_information https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/implicatures https://github.com/google/BIG-bench/tree/main/bigbench/benchmark_tasks/winowhy https://www.deepmind.com/publications/an-empirical-analysis-of-compute-optimal-large-language-model-training https://lambdalabs.com/blog/nvidia-h100-gpu-deep-learning-performance-analysis#:~:text=Compared%20to%20NVIDIA's%20previous%2Dgeneration,multiprocessors%2C%20and%20higher%20clock%20frequency. https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 27 min

    The Two Best AI Models/Enemies Just Got Released Simultaneously

    The two models that you will hear discussed for at least the next two months - Claude Opus 4.6 and GPT 5.3 Codex - just got released within 26 mins or each other. The full breakdown of around 250 pages of reports, with just the most interest moments, from the battle of which is best, Claude personhood, the surprising misbehaviour of Opus 4.6, and much more https://assemblyai.com/aiexplained Check out my fast-growing (!) app, free to use, and code INSIDER15 for Pro: https://lmcouncil.ai AI Insiders ($9): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 00:54 - Self-improvement? 02:44 - Knowledge Work 05:30 - Overly agentic behaviour 09:12 - Who Shouldn’t Use Claude Opus 11:39 - Step-change? 15:09 - Claude’s ‘Personhood’ Hassabis Roadmap: https://www.patreon.com/posts/hassabis-roadmap-149750869 Release of Opus 4.6: https://www.anthropic.com/news/claude-opus-4-6 212 Page System Card: https://www-cdn.anthropic.com/0dd865075ad3132672ee0ab40b05a53f14cf5288.pdf Claude Code Tip: https://x.com/bcherny/status/2019475897691124107 GPT Codex 5.3: https://openai.com/index/introducing-gpt-5-3-codex/ System Card: https://openai.com/index/gpt-5-3-codex-system-card/ Browse Comp: https://arxiv.org/pdf/2504.12516v1 Finance Agent: https://www.vals.ai/benchmarks/finance_agent Terminal Bench 2: https://arxiv.org/pdf/2601.11868 Vending Bench: https://andonlabs.com/blog/opus-4-6-vending-bench My X post: https://x.com/AIExplainedYT/status/2016851303436095647 Anthropic Apology: https://x.com/ch402/status/2014066134194995256/photo/1 Altman rebuttal: https://x.com/sama/status/2019139174339928189 https://x.com/sama/status/2019140276246442089 4% of GitHub: https://x.com/dylan522p/status/2019490550911766763 Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 24 min

    Orca: The Model Few Saw Coming

    The first model set to be opensourced that actually comes close to ChatGPT, and is just 13B (that's small enough for a laptop). The 51 page report from Microsoft was released just 48 hours ago but I have gone through it all, and bring relevant insights from 5 other papers. By imitating the logic and explanations of GPT 4 (and using GPT 3.5 as an assistant), as well as by training on diverse tasks and an order of magnitude more examples, we have Orca. I will showcase it on a dozen benchmarks and go through in detail how it works and why. I will also end on comments from Sam Altman and Ilya Sutskever on whether Opensource will catch-up... Orca Paper: https://arxiv.org/pdf/2306.02707.pdf False Promise Paper: https://arxiv.org/pdf/2305.15717.pdf https://sharegpt.com/ FLAN: https://arxiv.org/pdf/2301.13688.pdf Vicuna: https://lmsys.org/blog/2023-03-30-vicuna/ No Moat memo: https://www.semianalysis.com/p/google-we-have-no-moat-and-neither LLM Leaderboard: https://chat.lmsys.org/?leaderboard AGIEval: https://arxiv.org/pdf/2304.06364.pdf BIG-Bench Hard: https://arxiv.org/pdf/2210.09261.pdf Language Models as Tool Makers: https://arxiv.org/pdf/2305.17126.pdf Altman Interview: https://www.youtube.com/watch?v=VWUhASix9ws DERA Paper: https://arxiv.org/abs/2303.17071 Let's Verify Step by Step: https://arxiv.org/abs/2305.20050 https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 28 min

    You Are Being Told Contradictory Things About AI

    With headlines of an imminent job apocalypse, code red for ChatGPT and recursive self-improvement, at the same time as Anthropic's CEO yesterday saying we know how to scale to AGI, and Gemini 3 DeepThink out today, it is easy to get lost among the narratives and counter-narratives. So here are both, plus the facts behind them, for you to decide. https://epoch.ai/ai-explained-datacenters Epoch AI is the sponsor of today’s video, and my views, and those expressed in this video, do not necessarily reflect Epoch AI’s views in any way. Chapters: 00:00 - Introduction 00:42 - Job Apocalypse? 01:45 - Scaling to AGI 04:15 - Recursive Self-Improvement Needed, or Not 09:57 - OpenAI Code Red vs Gemini 3 DeepThink vs Claude Opus 4.5 13:27 - DeepSeek Speciale vs Mistral Large v3 16:45 - Claude Soul Document https://lmcouncil.ai/ AI Insiders ($9!): https://www.patreon.com/AIExplained Guardian Interview: https://www.theguardian.com/technology/ng-interactive/2025/dec/02/jared-kaplan-artificial-intelligence-train-itself MIT Study on Jobs/Tasks: https://iceberg.mit.edu/report.pdf vs https://www.cnbc.com/2025/11/26/mit-study-finds-ai-can-already-replace-11point7percent-of-us-workforce.html Amodei on Scaling: https://www.youtube.com/watch?v=FEj7wAjwQIk Claude Soul Document: https://www.lesswrong.com/posts/vpNG99GhbBoLov9og/claude-4-5-opus-soul-document Capabilities Original Stance: https://www.anthropic.com/news/core-views-on-ai-safety Ilya Interview: https://www.dwarkesh.com/p/ilya-sutskever-2 Ricursive Intelligence: https://x.com/RicursiveAI/status/1995932204703346946 Economist Worker Usage of GenAI: https://www.economist.com/finance-and-economics/2025/11/26/investors-expect-ai-use-to-soar-thats-not-happening#selection-1409.94-1413.42 Mistral v3 Large: https://docs.mistral.ai/models/mistral-large-3-25-12 Compute Slowdown Paper: https://joel-becker.com/images/publications/forecasting_time_horizon_under_compute_slowdown.pdf https://x.com/joel_bkr/status/1993023436541903155 METR Chart: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/ https://www.theinformation.com/articles/openais-350-billion-computing-cost-problem?rc=sy0ihq OpenAI Code Red: https://www.anthropic.com/news/core-views-on-ai-safety Rocket Company: https://www.independent.co.uk/news/world/americas/sam-altman-rocket-elon-musk-spacex-b2878351.html DeepSeek Paper: https://arxiv.org/html/2512.02556v1 DeepSeek Crowdstrike CCP: https://www.crowdstrike.com/en-us/blog/crowdstrike-researchers-identify-hidden-vulnerabilities-ai-coded-software/ https://simple-bench.com/ Patreon Post: https://www.patreon.com/c/aiexplained/posts Robot: https://x.com/jloganolson/status/1985850115379351799 Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 29 min

    Gemini 3 Pro: Breakdown

    Gemini 3 Pro is out, and records fell like snowflakes in Svalbard. No long description, chapters or links today, huge technical difficulties, including with audio, so just want to publish asap. https://app.grayswan.ai/ai-explained https://lmcouncil.ai AI Insiders ($9!): https://www.patreon.com/AIExplained Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 22 min

    AI Declarations and AGI Timelines – Looking More Optimistic?

    A head-spinning week of declarations, timelines, updates, papers and debates. I will cover as much as I can, not least the new all-in one interface, EmotionPrompt, executive orders analysed, Bletchley declaration, a new MIT safety paper, responsible scaling, LLMs for Chip Design, and a guest appearance from Dyson Spheres. https://www.patreon.com/AIExplained Gates Interview: https://www.handelsblatt.com/technik/ki/bill-gates-mit-ki-koennen-medikamente-viel-schneller-entwickelt-werden/29450298.html Dwarkesh Interview Legg: https://www.youtube.com/watch?v=Kc1atfJkiJU Dwarkesh Interview Christiano: https://www.youtube.com/watch?v=9AAhTLa0dT0 Representation Engineering: https://arxiv.org/pdf/2310.01405.pdf MLC: https://www.scientificamerican.com/article/new-training-method-helps-ai-generalize-like-people-do/ Anthropic Policy: https://www.anthropic.com/index/uk-ai-safety-summit Science Automation: https://arxiv.org/ftp/arxiv/papers/2304/2304.05332.pdf https://arxiv.org/pdf/2304.05376.pdf Jim Fan Tweet: https://twitter.com/DrJimFan/status/1719733318521086332 White House Executive Order: https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/ ChatGPT Update: https://twitter.com/NorthstarBrain/status/1718708111706636403?s=20 RSP policies by AGI Labs: https://www.aisafetysummit.gov.uk/policy-updates/#company-policies YouGov Poll: https://d3nkl3psvxxpe9.cloudfront.net/documents/ControlAI_AI_231019.pdf https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 12 min

    How Well Can GPT-4 See? And the 5 Upgrades That Are Next

    What exactly can GPT-4 see? And does OpenAI have a plan for physical embodiment? Over 12 separate developments announced in the last few days are covered in this video, in which I showcase how improvements in text, speech, 3D, and embodiment are beginning to complement each other. We'll take an early look at use cases for images in GPT 4, how Nvidia and Adobe are pushing boundaries in Text to 3D, at Sam Altman's Moore's Law For Everything, at the new LeRF model and Assembly AI's latest Speech-to-Text. We will also see how OpenAI already had a Robotics team but now is invested in 1x, developing humanoid robots. But they are not the first, as Agility Robotics has shown... Conformer Model: https://www.assemblyai.com/playground/source Napkin-to-Code: https://twitter.com/rowancheung/status/1635744529587359756 BeMyEyes: https://twitter.com/BeMyEyes/status/1635690254689599488 Tanishq: https://twitter.com/iScienceLuvr Medical Questions: https://arxiv.org/pdf/2303.13375.pdf VQA: https://arxiv.org/pdf/1904.08920.pdf Scale is All You Need: https://twitter.com/markchen90/status/1637903780393160704 Nvidia Text-to-3D: https://twitter.com/chenhsuan_lin/status/1638543416723406849?t=rXEsDcEvsdmMPEsZPPhWmw&s=09 Adobe Firefly: https://www.youtube.com/watch?v=_sJfNfMAQHw Moore's Law For Everything: https://moores.samaltman.com/#:~:text=Moore%27s%20Law%20for%20Everything.%20Broadly%20speaking%2C%20there%20are,we%20can%20get%20with%20the%20resources%20we%20have. LeRF 3D: https://arxiv.org/pdf/2303.09553.pdf Wired: https://www.wired.com/story/fast-forward-for-smarter-robots-just-add-humans/ Conformer Model: https://www.assemblyai.com/playground/source https://www.assemblyai.com/blog/conformer-1/ Boston Dynamics: https://www.youtube.com/watch?v=-e1_QhJ1EhQ 1X: https://www.1x.tech/ Quadrupedal Agility: https://arxiv.org/pdf/2303.11330.pdf Rubik's Cube: https://openai.com/research/solving-rubiks-cube Statues: https://twitter.com/TuurDemeester/status/1639421727548616710 GPT 4 Technical Report: https://arxiv.org/pdf/2303.08774.pdf Agility Robotics: https://www.youtube.com/watch?v=FhlUI2xDxdE Sam Altman Lex Fridman: https://www.youtube.com/watch?v=L_Guz73e6fw https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 20 min

    o3 breaks (some) records, but AI becomes pay-to-win

    A green card, o3 vs Gemini 2.5, 6 Benchmarks and a whole bunch of my thoughts on what on earth is happening in AI, from here to 2030. Plus, how AI is becoming pay-to-win, and why. Crazy times, 14 mins probably wasn’t enough. https://app.grayswan.ai/ai-explained AI Insiders ($9!): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 00:33 - FictionLiveBench 01:37 - PHYBench 02:14 - SimpleBench 02:54 - Virology Capabilities Test 03:13 - Mathematics Performance 04:29 - Vision Benchmarks 05:43 - V* and how o3 works 06:44 - Revenue and costs for you 08:54 - Expensive RL and trade-offs 09:40 - How to spend the OOMs 13:27 - Gray Swan Arena Green Card: https://techcrunch.com/2025/04/25/an-openai-researcher-who-worked-on-gpt-4-5-had-their-green-card-denied/ PHYBench: https://arxiv.org/pdf/2504.16074Virologytest: https://www.virologytest.ai/ How o3 Vision Works: https://arxiv.org/pdf/2312.14135 https://x.com/sainingxie/status/1912570624523829573 Visual puzzles: https://neulab.github.io/VisualPuzzles/ Fiction Bench: https://x.com/ficlive/status/1912863028141244850 https://geobench.org/ https://simple-bench.com/ AIME 2025: https://openai.com/index/introducing-o3-and-o4-mini/ USAMO: https://x.com/mbalunovic/status/1914398518896193747 NaturalBench: https://linzhiqiu.github.io/papers/naturalbench/ Where’s Waldo: https://uk.pinterest.com/pin/492792384225896298/ IMO and AlphaProof:https://deepmind.google/discover/blog/ai-solves-imo-problems-at-silver-medal-level/ Crazy Revenue: https://www.theinformation.com/articles/openai-forecasts-revenue-topping-125-billion-2029-agents-new-products-gain?rc=sy0ihq Number of Users: https://www.theinformation.com/briefings/googles-gemini-user-numbers-revealed-court?rc=sy0ihq Subscriptions pay to win: https://www.forbes.com/sites/paulmonckton/2025/04/23/google-leak-reveals-new-gemini-ai-subscription-levels/ GPU Trade-offs: https://x.com/sama/status/1915098951067554030 RL Scale-up Amodei: https://www.darioamodei.com/post/on-deepseek-and-export-controls Log-linear Returns: https://x.com/bobmcgrewai/status/1895228291981943265 2030 Scaling: https://epoch.ai/blog/can-ai-scaling-continue-through-2030 Model Size: https://x.com/slow_developer/status/1874554473256997201 Adam on AGI: https://x.com/TheRealAdamG/status/1913998366632968381 Papers on Patreon: https://arxiv.org/pdf/2502.01839 https://arxiv.org/pdf/2504.13837 Chollet Quote: https://x.com/fchollet/status/1912934762580447447 OpenSim: https://opensim.stanford.edu/ Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 35 min

    When Will AI Models Blackmail You, and Why?

    In the last few days Anthropic have released an impressive honest account of how all models blackmail, no matter what goal they have, and despite prompt warnings, and other preventions. But do these models *want* this? Thanks to Storyblocks for sponsoring this video! Download unlimited stock media at one set price with Storyblocks: https://storyblocks.com/AIExplained AI Insiders ($9!): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 01:20 - What prompts blackmail? 02:44 - Blackmail walkthrough 06:04 - ‘American interests’ 08:00 - Inherent desire? 10:45 - Switching Goals 11:35 - Murder 12:22 - Realizing it’s a scenario? 15:02 - Prompt engineering fix? 16:27 - Any fixes? 17:45 - Chekov’s Gun 19:25 - Job implications 21:19 - Bonus Details Report: https://www.anthropic.com/research/agentic-misalignment 30 Page Appendices: https://assets.anthropic.com/m/6d46dac66e1a132a/original/Agentic_Misalignment_Appendix.pdf Announcement: https://x.com/AnthropicAI/status/1936144602446082431?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Etweet OpenAI Files: https://www.openaifiles.org/ Grok 4 News: https://x.com/RonFilipkowski/status/1936372579607912473 Claude 4 Report Card: https://www-cdn.anthropic.com/6be99a52cb68eb70eb9572b4cafad13df32ed995.pdf New Apollo Research: https://www.apolloresearch.ai/blog/more-capable-models-are-better-at-in-context-scheming Interesting Reflections: https://nostalgebraist.tumblr.com/post/785766737747574784/the-void Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 30 min

    New Claude Opus 4.8: 15 Things You May’ve Missed

    The ‘best’ generally available AI model just dropped, but there is plenty I bet you missed about what it is, how it performs, and what the release tells us. 15 highlights from the 244 page system card, plus private testing, leader interview and more. AI Insiders ($9!): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 00:49 - Mythos in Weeks 01:49 - Adaptive not necessary 02:26 - Honesty? 04:37 - Flagging Uncertainty 04:57 - Benchmarks 08:54 - Mythos will be even better 10:30 - Business skillz 11:15 - Model Welfare 12:16 - Cyber Comparable 13:10 - Misalignment Concerns 16:22 - Meta Inabilities 17:58 - Code flagging 18:34 - Go to sleep 18:50 - Fast Mode 20:21 - Dynamic Workflows Opus 4.8 Paper: https://cdn.sanity.io/files/4zrzovbb/website/c886650a2e96fc0925c805a1a7ca77314ccbf4a6.pdf Release: https://www.anthropic.com/news/claude-opus-4-8 Chips: https://www.theinformation.com/articles/anthropic-talks-use-microsofts-ai-chips?rc=sy0ihq https://www.anthropic.com/news/expanding-our-use-of-google-cloud-tpus-and-services https://www.anthropic.com/news/higher-limits-spacex Patreon Vid: https://www.patreon.com/posts/re-up-anthropics-159289449 GDPVal: https://artificialanalysis.ai/evaluations/omniscience https://arxiv.org/abs/2510.04374 Amodei Technical Debt: https://www.youtube.com/watch?v=7xco5Qd2Oo8 Dynamic Workflows: https://x.com/ClaudeDevs/status/2060044853279617150 https://x.com/_catwu/status/2060054180379689074/photo/1 https://claude.com/blog/introducing-dynamic-workflows-in-claude-code https://simple-bench.com/ Check out my fast-growing (!) app, free to use, and code INSIDER15 for paid tiers: https://lmcouncil.ai Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 17 min

    What's Up With Bard? 9 Examples + 6 Reasons Google Fell Behind [ft. Muse, Med-PaLM 2 and more]

    A definitive comparison between Bard and GPT-4 - and it doesn’t look good for Bard. I demonstrate, with nine different examples, that Bard has a serious problem. From coding to content creation, summarisation to self-tutoring, I cover all the deficiencies that will stop Bard being currently useful. I then go into 6 explanations for why Google might have fallen behind. From Attention authors leaving the ship, not defining their product, safety and Anthropic, better models and Med-PaLM, I explore all the reasons why Bard isn’t up to scratch. Even pdf scanning doesn’t work. Dumpster Fire Article: https://www.cnbc.com/2023/02/10/google-employees-slam-ceo-sundar-pichai-for-rushed-bard-announcement.html Bard: https://bard.google.com/ GPT-4: https://chat.openai.com/chat?model=gpt-4 NYT Article: https://www.nytimes.com/2023/03/22/business/stock-markets-rates-fed-decision.html 8 authors leaving: https://www.forbes.com/sites/richardnieva/2023/02/08/google-openai-chatgpt-microsoft-bing-ai/?sh=648e593e4de4 Product Lead Comments: https://www.cnbc.com/2023/03/03/google-execs-say-in-all-hands-meeting-bard-ai-isnt-all-for-search-.html Anthropic Investment: https://www.theverge.com/2023/2/3/23584540/google-anthropic-investment-300-million-openai-chatgpt-rival-claude Imagen Explanation: https://imagen.research.google/ Imagen Paper: https://arxiv.org/pdf/2205.11487.pdf Muse Paper: https://arxiv.org/pdf/2301.00704.pdf Med-PaLM: https://arxiv.org/abs/2212.13138 Med-PaLM 2: https://www.youtube.com/watch?v=TSEXVjr6m8w Physics: https://www.bbc.co.uk/bitesize/guides/zqs47p3/revision/5 Data Advantage: https://www.forbes.com/sites/richardnieva/2023/02/08/google-openai-chatgpt-microsoft-bing-ai/?sh=648e593e4de4 Virtuous Cycle: https://www.cnbc.com/2023/03/03/google-execs-say-in-all-hands-meeting-bard-ai-isnt-all-for-search-.html https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 23 min

    Google Takes No Prisoners Amid Torrent of AI Announcements

    Google just announced at least 12 things that are each worthy of a video, but here are the top I/O highlights. From Veo 3 to Deep Research now being useable, Deep Think breaking records to Gemini Diffusion, Gemini 2.5 Flash changing how AI is priced and GemmaVerse, SynthID Detector and Imagen 4. And even this intro is missing other announcements covered in the vid! And yes, they’ll be plenty of Veo 3 clips to enjoy… https://80000hours.org/aiexplained AI Insiders ($9!): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 00:48 - Veo 3 02:10 - Gemini 2.5 Flash 03:13 - Universal Assistant 03:47 - Usage Skyrockets + OpenAI dig 04:51 - Gemini Pro Deep Think 06:21 - Overviews and AI Mode 07:26 - Deep Research Updates (new) + Jules 08:53 - Make and Deploy Apps with Gemini 09:12 - Imagen 4 10:00 - Gemini Diffusion 11:46 - Try It On 12:17 - SynthID Detector 13:30 - GemmaVerse, SignGemma, Gemma3n, medGemma 14:24 - Outro + Clips Event: https://www.youtube.com/watch?v=o8NiE3XMPrM Ntaive Audio: https://aistudio.google.com/generate-speech Gemini Diffusion: https://deepmind.google/models/gemini-diffusion/#capabilities New Gemini 2.5 Flash: https://deepmind.google/models/gemini/flash/ SignGemma (See end of this vid): https://www.youtube.com/watch?v=GjvgtwSOCao Deep Think: https://blog.google/technology/google-deepmind/google-gemini-updates-io-2025/#flash-improvements Google Parallel Sampling: https://www.patreon.com/posts/next-level-good-127441188 Price Plans: https://blog.google/products/google-one/google-ai-ultra/ Imagen 4 Benchmarks: https://deepmind.google/models/imagen/ Jules: https://jules.google/ SynthID Detector: https://blog.google/technology/ai/google-synthid-ai-content-detector/ Veo 3 Benchmarks: https://deepmind.google/models/veo/evals/ MedGemma: https://deepmind.google/models/gemma/medgemma/ Build Apps: https://aistudio.google.com/apps Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 19 min

    Deadline Day for Autonomous AI Weapons and Mass Surveillance

    Will Anthropic be forced to make a version of Claude for war? And does a new paper expose the risks of Claude agents, in both OpenClaw and the field of war? Plus, 5 more twists in the story of the Pentagon versus Anthropic + some AI lab employees, and a petition that could change everything, or nothing... Check out my fast-growing (!) app, free to use, and code INSIDER15 for paid tiers: https://lmcouncil.ai AI Insiders ($9!): https://www.patreon.com/AIExplained Chapters: 00:00 - Introduction 00:44 - Deadline Day + Petition 02:42 - Twist 1: Existing Deal 03:26 - Twist 2: Existing Policy 04:21 - Twist 3: Twin Threats 05:54 - Twist 4: Interesting Objections 11:32 - Twist 5: Anthropic’s Dropped Policy Dario Statement: https://www.anthropic.com/news/statement-department-of-war Google/OpenAI Petition: https://notdivided.org/ Axios on Amodei Rejection: https://www.axios.com/2026/02/26/anthropic-rejects-pentagon-ai-terms FT on US Threat: https://www.ft.com/content/11d27612-d6c5-4cf7-94dd-f65603549b7f Politico on Latest: https://archive.ph/20260227013117/https://www.politico.com/news/2026/02/26/incoherent-hegseths-anthropic-ultimatum-confounds-ai-policymakers-00800135 The Verge on Current Deal: https://www.theverge.com/ai-artificial-intelligence/883456/anthropic-pentagon-department-of-defense-negotiations Anthropic RSP change: https://www.anthropic.com/news/responsible-scaling-policy-v3 Time Magazine on RSP: https://time.com/7380854/exclusive-anthropic-drops-flagship-safety-pledge/ Agent of Chaos Paper: https://x.com/NatalieShapira/status/2026062499599319526 AI Agent Reliability Paper: https://arxiv.org/pdf/2602.16666 My Patreon Video: https://www.patreon.com/posts/real-mystery-ai-151647211 Patreon Documentary: https://www.patreon.com/posts/our-new-age-of-133960279 Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 30 min

    o3 - wow

    o3 isn’t one of the biggest developments in AI for 2+ years because it beats a particular benchmark. It is so because it demonstrates a reusable technique through which almost any benchmark could fall, and at short notice. I’ll cover all the highlights, benchmarks broken, and what comes next. Plus, the costs OpenAI didn’t want us to know, Genesis, ARC-AGI 2, Gemini-Thinking, and much more. AI Insiders ($9!): https://www.patreon.com/AIExplained FrontierMath: https://epoch.ai/frontiermath https://arxiv.org/pdf/2411.04872 Chollet Statement:https://arcprize.org/blog/oai-o3-pub-breakthrough MLC Paper: https://www.scientificamerican.com/article/new-training-method-helps-ai-generalize-like-people-do/?utm_campaign=socialflow&utm_source=twitter&utm_medium=social AlphaCode 2: https://storage.googleapis.com/deepmind-media/AlphaCode2/AlphaCode2_Tech_Report.pdf Human Performance on ARC-AGI: https://arxiv.org/pdf/2409.01374v1 Wei Tweet ‘3 months’:https://x.com/_jasonwei/status/1870184982007644614 Deliberative Alignment Paper: https://openai.com/index/deliberative-alignment/ Brown Safety Tweet: https://x.com/polynoamial/status/1870196476908834893 Swe-Bench Verified: https://openai.com/index/introducing-swe-bench-verified/ Amodei Prediction: https://x.com/OfirPress/status/1858567863788769518 David Dohan: 16 hours https://x.com/dmdohan/status/1870171404093796638 OpenAI Personal Writing: https://openai.com/index/learning-to-reason-with-llms/ https://simple-bench.com/ John Hallman Tweet: https://x.com/johnohallman/status/1870233375681945725 00:00 - Introduction 01:19 - What is o3? 03:18 - FrontierMath 05:15 - o4, o5 06:03 - GPQA 06:24 - Coding, Codeforces + SWE-verified, AlphaCode 2 08:13 - 1st Caveat 09:03 - Compositionality? 10:16 - SimpleBench? 13:11 - ARC-AGI, Chollet 20:25 - Safety Implicaitons AI Insiders: https://www.patreon.com/AIExplained Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 30 min

    AGI: (gets close), Humans: ‘Who Gets to Own it?’

    A 'frontier reasoning model' from just 1000 examples (s1). A $100B Musk bid for power. Gemini 2, Rand and warning from Amodei. Here’s 7 developments you may have missed but which I would argue help us understand how the next few years will play out. From labour vs capital to automating rival companies and countries, and from non-profit shenanigans to new mini-docs, there was just too much for me not to make a vid. GiveWell: https://www.givewell.org/charities/top-charities AI Insiders ($9!): https://www.patreon.com/AIExplained s1 Paper: https://arxiv.org/pdf/2501.19393 Musk Bid: https://www.wsj.com/tech/ai/musks-97-4-billion-openai-bid-piles-pressure-on-altman-f6749e6c?mod=hp_lead_pos1 Altman Reply: https://x.com/sama/status/1889059531625464090?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Etweet Google vs OpenAI: https://x.com/sama/status/1888703820596977684 RAND Study: https://www.rand.org/pubs/perspectives/PEA3691-4.html Dev Meetup: https://x.com/btibor91/status/1888976302621040852 Altman $100 Trillion: https://www.nytimes.com/2023/03/31/technology/sam-altman-open-ai-chatgpt.html Karpathy Vid: https://www.youtube.com/watch?v=7xTGNNLPyMI Amodei Warning: https://www.anthropic.com/news/paris-ai-summit Bengio Source: https://www.youtube.com/watch?v=6HDjVncL5Go Chapters: 00:00 - Intro 01:37 - AGI Inches Closer 04:26 - ‘Super-Exponential’ 05:58 - Musk Bid 07:34 - Luxury Goods and Land 09:05 - ‘Benefits All Humanity’ 12:52 - ‘National Security’ 14:21 - s1 20:33 - Final thoughts Non-hype Newsletter: https://signaltonoise.beehiiv.com/ Podcast: https://aiexplainedopodcast.buzzsprout.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 19 min

    9 of the Best Bing (GPT 4) Prompts

    Everyone knows by now how to prompt ChatGPT, but what about Bing? Take prompt engineering to a whole new level with these 9 game-changing Bing Chat prompts. Did you know you can get interviewed by Bing, time travel, force Bing to improve its output and so much more? These are the best 9 prompts that I could find, after analysing over 200 prompts and trying out dozens of examples personally. You might call them prompt hacks, but I prefer to think of them as simply creative exploration of Bing's capacities. Some prompts inspired by this post: https://github.com/f/awesome-chatgpt-prompts And by https://twitter.com/emollick https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

  • Yesterday · 25 min

    11 Major AI Developments: RT-2 to '100X GPT-4'

    There were 11 major developments in AI in just this week, from RT-2 – a significant step on the path to robotic AGI – to ‘100x GPT 4 in 18 months’, and from the uplifting news of TranscribeGlass and AI Barbie-heimer to dramatic revelations about OpenAI in the Atlantic. We’ll also glimpse Stable Beluga 2 – the ‘first true competitor’ to ChatGPT, based on the open source Llama 2, hear about Universal Jailbreaks, learn that OpenAI has surrendered on Text Detection, plus I’ll cover the highlights of the Senate hearing on AI. https://www.patreon.com/AIExplained Chapters: 0:18 – RT-2 2:46 – 100X GPT-4 3:57 – AI Video 4:29 – Altman Atlantic 8:41 – Jan Leike Interview 10:02 – Speech Transcription + Generation 11:07 – OpenAI Text Surrender 11:43 – Stable Beluga 2 12:51 – Universal Jailbreaks 14:19 – Senate testimony: Bio 16:59 – Senate Testimony: Security RT-2: https://www.deepmind.com/blog/rt-2-new-model-translates-vision-and-language-into-action RT-2 Paper: https://robotics-transformer2.github.io/assets/rt2.pdf RT-2 NYT: https://www.nytimes.com/2023/07/28/technology/google-robots-ai.html Suleyman Barron’s Interview: https://www.barrons.com/articles/ai-chatbot-siri-alexa-inflection-pi-fa1809f8 OpenAI Atlantic: https://www.theatlantic.com/magazine/archive/2023/09/sam-altman-openai-chatgpt-gpt-4/674764/ TranscribeGlass: https://twitter.com/igorsushko/status/1684645290132000768 Barbie-heimer: https://twitter.com/terhavlova/status/1684265329277403149 Runway Gen 2: https://twitter.com/runwayml/status/1666429706932043776 ElevenLabs: https://elevenlabs.io/speech-synthesis Jan Leike Interview: https://www.lesswrong.com/posts/bsNXqHgiDA6dAKNun/24-superalignment-with-jan-leike Stable Beluga 2: https://stability.ai/blog/stable-beluga-large-instruction-fine-tuned-models Open LLM Leaderboard: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard Fast Takeoff: https://www.alignmentforum.org/posts/YgNYA6pj2hPSDQiTE/distinguishing-definitions-of-takeoff#:~:text=associated%20with%20testing.-,Fast,case%20in%20Yudkowsky's%20foom%20scenario. OpenAI Text Detection Surrender: https://techcrunch.com/2023/07/25/openai-scuttles-ai-written-text-detector-over-low-rate-of-accuracy/ Universal Jailbreaks: https://www.nytimes.com/2023/07/27/business/ai-chatgpt-safety-research.html https://llm-attacks.org/ https://twitter.com/mustafasuleymn?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor Senate Hearing Amodei: https://www.youtube.com/watch?v=hm1zexCjELo Altman Question from Philip @ AI Explained in Melbourne: https://www.youtube.com/watch?v=7SMzkBKzsQs Wired Bio Expert: https://www.wired.com/story/have-a-nice-future-podcast-15/ https://www.patreon.com/AIExplained Non-Hype, Free Newsletter: https://signaltonoise.beehiiv.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

Showing 41–60 of 145 episodes