LessWrong posts by zvi

“Claude Fable 5 and Mythos 5: Capabilities” by Zvi

1 h 21 min · 19 jun 2026
aflevering “Claude Fable 5 and Mythos 5: Capabilities” by Zvi artwork

Beschrijving

Only three days after the release of Claude Fable 5, Anthropic was forced by the United States Government to make it unavailable, when a jailbreak was brought to its attention, rather than the previous situation of ‘yes obviously experts can jailbreak anything if they care enough’ and ‘yes obviously you can ask Fable to fix your code.’ Three days was enough time for many of us to learn to love Fable, and for us to dearly miss it now that it is gone. The world was briefly smarter, and now it is again stupider. At some point it will get smarter again, which will likely be within two weeks. This post is written as if Fable 5 is again available for public use, rather than trying to include a lot of qualifying clauses. It remains to be seen how this will play out, and this post does not attempt to cover that question. My previous release coverage of Fable covered the model card and then model welfare. Coverage of the government takedown of Fable starts here, and continues here and here. The Official Pitch The pitch is that Fable 5 is the best model [...] --- Outline: (01:08) The Official Pitch (04:06) Technical Details (04:31) The System Prompt and Jailbreak (06:45) Benchmarks (15:22) Other People's Benchmarks (21:08) The Classifiers Are Not Messing Around (22:53) The Classifiers Need Work (28:15) The Classifiers Have Consequences (29:18) First Hit Is Free (29:53) How Easily We Forget (30:46) Data Retention Is An Issue (31:15) Fable For The Win (36:15) Andrej Karpathy Is Impressed (37:54) Every Is Very Impressed (39:04) Other People Are Impressed (51:10) Know How To Tell a Fable (53:06) You Can Just Make Things (55:37) You Can Just Install Things (56:05) Good Personality (57:51) Fable Writes A Fable (01:06:04) Is That Code (01:08:32) Fable Crosses The Threshold (01:09:12) Man With A Plan (01:10:12) Less Impressed Assessments (01:13:39) Actively Negative Assessments (01:14:16) Coherence (01:15:27) Good Night And Good Luck (01:16:05) Curious Fable (01:16:23) I See You, Baby (01:16:40) We Finally Did It We Know How To Count Letters (01:17:46) That's Not My Style (01:20:12) The Lighter Side --- First published: June 19th, 2026 Source: https://www.lesswrong.com/posts/kMnobCQp9z2pSbzDB/claude-fable-5-and-mythos-5-capabilities [https://www.lesswrong.com/posts/kMnobCQp9z2pSbzDB/claude-fable-5-and-mythos-5-capabilities?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration] --- Narrated by TYPE III AUDIO [https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=lesswrong&utm_campaign=ai_narration]. --- Images from the article: Performance comparison table of Claude family models versus other AI models across multiple benchmarks. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hiuggoqq0wxtieup1ogs]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hiuggoqq0wxtieup1ogs ---------------------------------------- Table showing benchmark performance scores across different AI models and evaluation tests. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/j3lgvgssyqu7nv5n76nd]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/j3lgvgssyqu7nv5n76nd ---------------------------------------- Line graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mhl0big4oz6dfsdsusby]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mhl0big4oz6dfsdsusby ---------------------------------------- Graph showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/xa6vj8cndixwwxpp5cly]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/xa6vj8cndixwwxpp5cly ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/qzjin04zw1r2ttogtc5e]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/qzjin04zw1r2ttogtc5e ---------------------------------------- Table showing Toolathlon scores for Claude AI models across different pass rates and average turns. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/scexyz016idar9ltxxu2]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/scexyz016idar9ltxxu2 ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hlp3tgx0ya3dhabylkm5]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hlp3tgx0ya3dhabylkm5 ---------------------------------------- Line graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/czd85rgkmaer1kru9rgu]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/czd85rgkmaer1kru9rgu ---------------------------------------- Bar graph showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/lyqrcpikudbnrp6eem7f]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/lyqrcpikudbnrp6eem7f ---------------------------------------- Two bar charts comparing AI model performance scores and cost per intelligence index task. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/w6xjiegp4e69erzlkiku]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/w6xjiegp4e69erzlkiku ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mkkjjwkblu842g06axs5]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mkkjjwkblu842g06axs5 ---------------------------------------- Benchmark leaderboard showing AI model performance on ProofBench math proofs. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/oui6viwzrlnqgprwaafo]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/oui6viwzrlnqgprwaafo ---------------------------------------- Bar chart showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/enatrmfceqxqlknz3d3e]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/enatrmfceqxqlknz3d3e ---------------------------------------- Bar chart titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/djg0hgltcll8fm3hgrrw]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/djg0hgltcll8fm3hgrrw ---------------------------------------- Comic comparing appropriate versus inappropriate workplace comments about appearance. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/grxrljaqb76qrrsiodrr]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/grxrljaqb76qrrsiodrr ---------------------------------------- Section titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/i5cu29iaomnnjmgxydkf]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/i5cu29iaomnnjmgxydkf ---------------------------------------- Poem titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/ox6awbgcmv95kqewjthb]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/ox6awbgcmv95kqewjthb ---------------------------------------- Bar graph showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jsw6jcemsexxa8x86nzp]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jsw6jcemsexxa8x86nzp ---------------------------------------- Bar chart showing new terms coined by AI models with adoption rates. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/vvismkxidzlnzhvg3olz]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/vvismkxidzlnzhvg3olz ---------------------------------------- Table showing terms, coiners, and usage statistics for various phrases coined by Fable 5. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jqgxlmo4dfhjbeyhmllz]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jqgxlmo4dfhjbeyhmllz ---------------------------------------- Table showing GPT-5.5 terms with coiner and usage counts. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/x1ppnfcsb70xfjhr7p9m]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/x1ppnfcsb70xfjhr7p9m ---------------------------------------- A Twitter thread shows a conversation about wordplay. The first tweet asks [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jogylfvetz6ylgqhx0ee]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jogylfvetz6ylgqhx0ee ---------------------------------------- AI chatbot conversation about whether to walk or drive to a car wash 100 feet away. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pxfzumvynyawfotoa2ep]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pxfzumvynyawfotoa2ep ---------------------------------------- Ten pithy tweets styled after tracewoodgrains about education, internet culture, and institutional dynamics. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pprr4di0ni2x98s7ye95]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pprr4di0ni2x98s7ye95 ---------------------------------------- List of ten humorous tweets in Joe Weisenthal's style about economics and finance. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/c9orpg966p3n6bzscqn1]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/c9orpg966p3n6bzscqn1 ---------------------------------------- Ten pithy tweets in @TheZvi style about regulation, prediction markets, AI policy, pricing, testing, economics, education standards, technology contradictions, and consciousness debates. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/eflubmdzdzqeg4vql6zr]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/eflubmdzdzqeg4vql6zr ---------------------------------------- List of ten humorous observations about AI assistant limitations and human interactions. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/r1928pqzunjzfq6j9nhi]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/r1928pqzunjzfq6j9nhi ---------------------------------------- Person facepalming emoji with blonde hair and blue shirt. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/5nDxmAvZ9w5CPa9gR/rim2lomv33dh8hv9ujz8]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/5nDxmAvZ9w5CPa9gR/rim2lomv33dh8hv9ujz8 Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts [https://pocketcasts.com/], or another podcast app.

Reacties

0

Wees de eerste die een reactie plaatst

Meld je nu aan en word lid van de LessWrong posts by zvi community!

Probeer gratis

Probeer 14 dagen gratis

€ 9,99 / maand na proefperiode. · Elk moment opzegbaar.

  • Podcasts die je alleen op Podimo hoort
  • 20 uur luisterboeken / maand
  • Gratis podcasts

Alle afleveringen

250 afleveringen

aflevering “Monthly Roundup #43: June 2026” by Zvi artwork

“Monthly Roundup #43: June 2026” by Zvi

Your monthly hit of all the things that are fit to print without a better place to live. Today is election day here in New York City, so again a reminder that if you are a registered Democrat and live in NY-12 today is the final day to vote for Alex Bores for Congress, and as per my argument yesterday that this matters a lot for ensuring we have a sensible Congressional response to AI. RIP FiveThirtyEight ABC and Disney completely take down FiveThirtyEight and all its articles, after telling Nate Silver they would refuse to sell it to him at any price because Nate had criticized their management of the brand. Nate Silver took this opportunity to reminisce and tell some stories about the old website, and the reasons the path of not seeking revenue and working with an entity too big to care ultimately doomed them. ‘What a bunch of assholes,’ indeed. I can grudgingly accept this sort of thing when it maximizes profits and the amount is meaningful, but this is different. Jack: This sort of digital arson is so frustrating. Pretty sure Dante had a place in mind for rights-holders [...] --- Outline: (00:33) RIP FiveThirtyEight (01:31) RIP Books (02:18) Bad News (09:53) Good Advice (18:31) Opportunity Knocks (19:15) Lower Awareness (22:21) The New York Times Has Some Issues (22:47) Liar Liar (25:51) Conspiracy Theory (26:16) Good News, Everyone (26:31) For Your Entertainment (28:00) A Matter of Taste (35:21) Gamers Gonna Game Game Game Game Game (37:33) I Was Promised Flying Self-Driving Cars (38:38) Sports Go Sports (39:09) Government Working (42:03) Jones Act Watch (43:02) Humans Can Be Strategic (44:58) Variously Effective Altruism (46:14) Support Anti-Aging Research (47:25) The Lighter Side --- First published: June 23rd, 2026 Source: https://www.lesswrong.com/posts/Taa4zmSNtD5S99tJT/monthly-roundup-43-june-2026 [https://www.lesswrong.com/posts/Taa4zmSNtD5S99tJT/monthly-roundup-43-june-2026?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration] --- Narrated by TYPE III AUDIO [https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=lesswrong&utm_campaign=ai_narration]. --- Images from the article: Three-panel illustration showing child pouring water between containers in science experiment. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/el9dklitbviiggottumc]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/el9dklitbviiggottumc ---------------------------------------- Matrix showing relationship between worry and outcomes, with suffering levels indicated by color coding. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/vsmvborve3fbvwt6bhc6]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/vsmvborve3fbvwt6bhc6 ---------------------------------------- Decision matrix showing worry versus outcome combinations with advice phrases. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/koxe2ojxjlutyrdojabg]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/koxe2ojxjlutyrdojabg ---------------------------------------- Infographic showing four steps of mental health campaign paradox with illustrations and explanatory text. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/gjxsaxyr5fdn8uvakil6]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/gjxsaxyr5fdn8uvakil6 ---------------------------------------- Traditional Chinese temple complex beside modern white cubic building with pilotis. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/q2vy8qcnvb7kwsmgsxqu]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/q2vy8qcnvb7kwsmgsxqu ---------------------------------------- Pie chart titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/splnms08x2ckxlaidwss]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/splnms08x2ckxlaidwss ---------------------------------------- Email from Twilio SendGrid announcing automatic Pride theme application to email headers and footers, with opt-out available. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/qp5pxmw2d2xxiurb03my]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/qp5pxmw2d2xxiurb03my ---------------------------------------- Email announcement about adding a [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/ajpwru30xrehujl0letl]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/Taa4zmSNtD5S99tJT/ajpwru30xrehujl0letl Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts [https://pocketcasts.com/], or another podcast app.

Gisteren48 min
aflevering “GLM-5.2 Is The New Best Open Model” by Zvi artwork

“GLM-5.2 Is The New Best Open Model” by Zvi

GLM-5.2 arrived last week. It boasts excellent benchmarks and looks strong. Benchmarks here are a de facto ceiling of how good it is, not a point estimate. Essentially all other aspects of an open model like this, beyond speed and price, will almost always be worse than the numbers suggest. Still, impressive. It is definitely a large step up from GLM-5.1, and likely the strongest open model. GLM-5.2 is still substantially behind the absolute frontier, although plausibly on the cost-benefit Pareto frontier. It seems closer to the frontier than previous efforts, including probably closer than DeepSeek R1 was during the DeepSeek moment. This is the new ‘peak close behind’ moment. Its existence is a substantial updates to push back some of the ‘where are all the updates’ updates in the opposite direction over time. Purely in terms of core tasks that GLM-5.2 is capable of doing, and ignoring missing features and its inferior generalization, and ignoring that it is distilled from Claude, and ignoring the Mythos class of models, and marking purely from date of public release, you can make a case GLM-5.2 is somewhere between 4 months and 7 months behind the frontier [...] --- Outline: (02:01) Alex Bores For Congress In NY-12 (03:41) Signs of Life (05:05) The Benchmarks (09:02) GLM-5.2 Is Distilled From Claude (09:55) Positive Responses (16:00) Finding The Niche (17:30) Negative Reactions (20:05) Looking To The Future --- First published: June 22nd, 2026 Source: https://www.lesswrong.com/posts/reXkwJbB8GYdeuvDt/glm-5-2-is-the-new-best-open-model [https://www.lesswrong.com/posts/reXkwJbB8GYdeuvDt/glm-5-2-is-the-new-best-open-model?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration] --- Narrated by TYPE III AUDIO [https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=lesswrong&utm_campaign=ai_narration]. --- Images from the article: Graph showing DeepSWE score versus average cost per task for various AI models. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/reXkwJbB8GYdeuvDt/hw8vmlgx4xk0ubcworrm]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/reXkwJbB8GYdeuvDt/hw8vmlgx4xk0ubcworrm Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts [https://pocketcasts.com/], or another podcast app.

22 jun 202621 min
aflevering “Claude Fable 5 and Mythos 5: Capabilities” by Zvi artwork

“Claude Fable 5 and Mythos 5: Capabilities” by Zvi

Only three days after the release of Claude Fable 5, Anthropic was forced by the United States Government to make it unavailable, when a jailbreak was brought to its attention, rather than the previous situation of ‘yes obviously experts can jailbreak anything if they care enough’ and ‘yes obviously you can ask Fable to fix your code.’ Three days was enough time for many of us to learn to love Fable, and for us to dearly miss it now that it is gone. The world was briefly smarter, and now it is again stupider. At some point it will get smarter again, which will likely be within two weeks. This post is written as if Fable 5 is again available for public use, rather than trying to include a lot of qualifying clauses. It remains to be seen how this will play out, and this post does not attempt to cover that question. My previous release coverage of Fable covered the model card and then model welfare. Coverage of the government takedown of Fable starts here, and continues here and here. The Official Pitch The pitch is that Fable 5 is the best model [...] --- Outline: (01:08) The Official Pitch (04:06) Technical Details (04:31) The System Prompt and Jailbreak (06:45) Benchmarks (15:22) Other People's Benchmarks (21:08) The Classifiers Are Not Messing Around (22:53) The Classifiers Need Work (28:15) The Classifiers Have Consequences (29:18) First Hit Is Free (29:53) How Easily We Forget (30:46) Data Retention Is An Issue (31:15) Fable For The Win (36:15) Andrej Karpathy Is Impressed (37:54) Every Is Very Impressed (39:04) Other People Are Impressed (51:10) Know How To Tell a Fable (53:06) You Can Just Make Things (55:37) You Can Just Install Things (56:05) Good Personality (57:51) Fable Writes A Fable (01:06:04) Is That Code (01:08:32) Fable Crosses The Threshold (01:09:12) Man With A Plan (01:10:12) Less Impressed Assessments (01:13:39) Actively Negative Assessments (01:14:16) Coherence (01:15:27) Good Night And Good Luck (01:16:05) Curious Fable (01:16:23) I See You, Baby (01:16:40) We Finally Did It We Know How To Count Letters (01:17:46) That's Not My Style (01:20:12) The Lighter Side --- First published: June 19th, 2026 Source: https://www.lesswrong.com/posts/kMnobCQp9z2pSbzDB/claude-fable-5-and-mythos-5-capabilities [https://www.lesswrong.com/posts/kMnobCQp9z2pSbzDB/claude-fable-5-and-mythos-5-capabilities?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration] --- Narrated by TYPE III AUDIO [https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=lesswrong&utm_campaign=ai_narration]. --- Images from the article: Performance comparison table of Claude family models versus other AI models across multiple benchmarks. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hiuggoqq0wxtieup1ogs]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hiuggoqq0wxtieup1ogs ---------------------------------------- Table showing benchmark performance scores across different AI models and evaluation tests. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/j3lgvgssyqu7nv5n76nd]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/j3lgvgssyqu7nv5n76nd ---------------------------------------- Line graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mhl0big4oz6dfsdsusby]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mhl0big4oz6dfsdsusby ---------------------------------------- Graph showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/xa6vj8cndixwwxpp5cly]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/xa6vj8cndixwwxpp5cly ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/qzjin04zw1r2ttogtc5e]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/qzjin04zw1r2ttogtc5e ---------------------------------------- Table showing Toolathlon scores for Claude AI models across different pass rates and average turns. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/scexyz016idar9ltxxu2]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/scexyz016idar9ltxxu2 ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hlp3tgx0ya3dhabylkm5]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/hlp3tgx0ya3dhabylkm5 ---------------------------------------- Line graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/czd85rgkmaer1kru9rgu]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/czd85rgkmaer1kru9rgu ---------------------------------------- Bar graph showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/lyqrcpikudbnrp6eem7f]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/lyqrcpikudbnrp6eem7f ---------------------------------------- Two bar charts comparing AI model performance scores and cost per intelligence index task. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/w6xjiegp4e69erzlkiku]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/w6xjiegp4e69erzlkiku ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mkkjjwkblu842g06axs5]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/mkkjjwkblu842g06axs5 ---------------------------------------- Benchmark leaderboard showing AI model performance on ProofBench math proofs. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/oui6viwzrlnqgprwaafo]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/oui6viwzrlnqgprwaafo ---------------------------------------- Bar chart showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/enatrmfceqxqlknz3d3e]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/enatrmfceqxqlknz3d3e ---------------------------------------- Bar chart titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/djg0hgltcll8fm3hgrrw]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/djg0hgltcll8fm3hgrrw ---------------------------------------- Comic comparing appropriate versus inappropriate workplace comments about appearance. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/grxrljaqb76qrrsiodrr]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/grxrljaqb76qrrsiodrr ---------------------------------------- Section titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/i5cu29iaomnnjmgxydkf]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/i5cu29iaomnnjmgxydkf ---------------------------------------- Poem titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/ox6awbgcmv95kqewjthb]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/ox6awbgcmv95kqewjthb ---------------------------------------- Bar graph showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jsw6jcemsexxa8x86nzp]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jsw6jcemsexxa8x86nzp ---------------------------------------- Bar chart showing new terms coined by AI models with adoption rates. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/vvismkxidzlnzhvg3olz]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/vvismkxidzlnzhvg3olz ---------------------------------------- Table showing terms, coiners, and usage statistics for various phrases coined by Fable 5. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jqgxlmo4dfhjbeyhmllz]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jqgxlmo4dfhjbeyhmllz ---------------------------------------- Table showing GPT-5.5 terms with coiner and usage counts. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/x1ppnfcsb70xfjhr7p9m]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/x1ppnfcsb70xfjhr7p9m ---------------------------------------- A Twitter thread shows a conversation about wordplay. The first tweet asks [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jogylfvetz6ylgqhx0ee]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/jogylfvetz6ylgqhx0ee ---------------------------------------- AI chatbot conversation about whether to walk or drive to a car wash 100 feet away. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pxfzumvynyawfotoa2ep]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pxfzumvynyawfotoa2ep ---------------------------------------- Ten pithy tweets styled after tracewoodgrains about education, internet culture, and institutional dynamics. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pprr4di0ni2x98s7ye95]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/pprr4di0ni2x98s7ye95 ---------------------------------------- List of ten humorous tweets in Joe Weisenthal's style about economics and finance. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/c9orpg966p3n6bzscqn1]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/c9orpg966p3n6bzscqn1 ---------------------------------------- Ten pithy tweets in @TheZvi style about regulation, prediction markets, AI policy, pricing, testing, economics, education standards, technology contradictions, and consciousness debates. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/eflubmdzdzqeg4vql6zr]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/eflubmdzdzqeg4vql6zr ---------------------------------------- List of ten humorous observations about AI assistant limitations and human interactions. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/r1928pqzunjzfq6j9nhi]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/kMnobCQp9z2pSbzDB/r1928pqzunjzfq6j9nhi ---------------------------------------- Person facepalming emoji with blonde hair and blue shirt. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/5nDxmAvZ9w5CPa9gR/rim2lomv33dh8hv9ujz8]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/5nDxmAvZ9w5CPa9gR/rim2lomv33dh8hv9ujz8 Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts [https://pocketcasts.com/], or another podcast app.

19 jun 20261 h 21 min
aflevering “AI #173: AI Pauses” by Zvi artwork

“AI #173: AI Pauses” by Zvi

A lot of things are always happening. Only one story matters. Claude Fable 5 and Claude Mythos 5 were shut down, by the White House, via an imposition of export controls at 5:23pm on Friday, wreaking all sorts of havoc. There was then a scramble. Anthropic flew its people out to Washington, where they met with the Trump Administration on Monday, with hopes expressed that this could be quickly resolved. What caused this? The Trump Administration said it was due to a jailbreak of Fable, which we now know they were told about by Amazon. They called Dario Amodei, who they complain did not take the issue sufficiently seriously. Rather than shutting down the model, he tried to explain why he saw no need to do that. This did not go well. The ‘jailbreak’ turns out to be saying ‘fix this code,’ and the demo was getting Fable to find the same weaknesses that were easily identified by Opus 4.8 and GPT-5.5. As in, Fable is willing to work to fix security vulnerabilities if you give it a codebase. From this information and process, you could then figure out what the original bug in the [...] --- Outline: (02:40) Language Models Offer Mundane Utility (02:51) Language Models Don't Offer Mundane Utility (03:14) Huh, Upgrades (03:44) On Your Marks (08:43) VirtueBench (10:40) Choose Your Fighter (11:20) Papers, Please (11:48) Deepfaketown and Botpocalypse Soon (13:32) Goodhart's Law Strikes Again (14:23) They Took Our Jobs (16:49) The MidJourney Full Body Imaging Scanner (19:16) Introducing (20:36) In Other AI News (22:47) Show Me the Money (23:18) Bubble, Bubble, Toil and Trouble (24:51) Quiet Speculations (27:15) People Just Say Things (30:30) The Widened Path (32:34) Scott Alexander Lays Out His AI Opinions (38:36) Quickly, There's No Time (39:50) Policy On The AI Exponential (49:36) Anthropic Offers Two Policy Frameworks (50:46) Obligations of Developers (55:11) Societal Resilience Measures (56:20) Economic Policy Framework (01:01:26) White House Pauses AI Deployment (01:10:14) The Once And Future Fable (01:15:29) How To Fix This Code (01:17:14) The End of Privacy (01:18:45) AIs Have Preferences (01:20:56) The Quest for Sane Regulations (01:23:37) Chip City (01:24:14) The Week in Audio (01:24:25) Rhetorical Innovation (01:25:03) Aligning a Smarter Than Human Intelligence is Difficult (01:26:40) People Are Worried About AI Killing Everyone (01:27:53) The Lighter Side The original text contained 2 footnotes which were omitted from this narration. --- First published: June 18th, 2026 Source: https://www.lesswrong.com/posts/P7jBmCeBDq2ebojWY/ai-173-ai-pauses [https://www.lesswrong.com/posts/P7jBmCeBDq2ebojWY/ai-173-ai-pauses?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration] --- Narrated by TYPE III AUDIO [https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=lesswrong&utm_campaign=ai_narration]. --- Images from the article: Bar chart comparing AI model performance to human baseline on benchmark score. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/pkyfp4pk6z4lqwgs5pvn]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/pkyfp4pk6z4lqwgs5pvn ---------------------------------------- Bar charts showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/aqwlaj7a8ofvvghvxdnn]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/aqwlaj7a8ofvvghvxdnn ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/p1iblqa341j8pfzdaa6y]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/p1iblqa341j8pfzdaa6y ---------------------------------------- News article screenshot. The headline reads: [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/hqqhjbiklqkrknia7lwl]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/hqqhjbiklqkrknia7lwl ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/rbelsstadenxlcx4g028]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/rbelsstadenxlcx4g028 ---------------------------------------- Bar graph titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/quud8mcjjzwyav9zvrd4]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/quud8mcjjzwyav9zvrd4 ---------------------------------------- Bar chart showing nine modes of work in Claude Code sessions by percentage share. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/ewtopss1qdz369qfqn5r]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/ewtopss1qdz369qfqn5r ---------------------------------------- Bar graph showing actions and output words per user expertise level with statistical annotations. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/bfddlkcldj1ve7ze9fii]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/bfddlkcldj1ve7ze9fii ---------------------------------------- Two line graphs showing composition and value changes of Claude Code work from October 2025 to April 2026. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/rrrpy0h2cekyq8e1jbrv]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/rrrpy0h2cekyq8e1jbrv ---------------------------------------- Meme format showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/myn4acdfcoqtlsw0hlqg]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/myn4acdfcoqtlsw0hlqg ---------------------------------------- David Sacks tweets: [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/joqsctwuxmptrknaalob]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/joqsctwuxmptrknaalob ---------------------------------------- OpenAI and Anthropic bosses join G7 leaders for AI lunch in France. The image shows two men at a table during what appears to be a formal meeting, with microphones and water bottles visible. The left person wears a suit and the right person wears a red tie. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/w1dan9qjtk9txfxjpgd4]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/w1dan9qjtk9txfxjpgd4 ---------------------------------------- Tier ranking of tech industry leaders from S to D tier. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/qil9yricatgmjicslger]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/qil9yricatgmjicslger ---------------------------------------- Tier list ranking tech companies: Anthropic, DeepMind, AMD, Microsoft, Stripe, Nvidia, Intel in S tier; IBM, Apple, Databricks, SpaceX, Salesforce, Google, Adobe, Tesla, OpenAI, Airbnb, Netflix in A tier; Amazon, DeepSeek, Oracle in B tier; Uber, Robinhood, Coinbase in C tier; xAI, Meta, Palantir in D tier. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/blyr73ezqy13v8ht0ab5]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/blyr73ezqy13v8ht0ab5 ---------------------------------------- Country flags organized in a tier list from S to D ranks. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/djmk3nvnlynrxh6jodxm]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/djmk3nvnlynrxh6jodxm ---------------------------------------- Bar graph showing cheating rates across human time buckets for Time Horizon tasks. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/cpw5vvbffdmzdakv76pl]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/cpw5vvbffdmzdakv76pl ---------------------------------------- Screenshot showing [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/xrw0c7qekiczj4f0nwgw]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/xrw0c7qekiczj4f0nwgw ---------------------------------------- Screenshot of text titled [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/rfbczhv7w9vhowgcje5m]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/rfbczhv7w9vhowgcje5m ---------------------------------------- Gary Marcus tweets: [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/papshhnaty4m2xb5davk]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/papshhnaty4m2xb5davk ---------------------------------------- Eliezer Yudkowsky tweets: [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/xhctpdffsnidhi0zrhle]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/P7jBmCeBDq2ebojWY/xhctpdffsnidhi0zrhle Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts [https://pocketcasts.com/], or another podcast app.

18 jun 20261 h 33 min
aflevering “The Once And Future Fable #3: Fix This Code” by Zvi artwork

“The Once And Future Fable #3: Fix This Code” by Zvi

The mainstream media continues to sleep on the most important story in the world. It has now been two days since Anthropic flew its people out to Washington, and I offered my previous update. We have heard nothing back from those meetings. Prediction market prices have moved rapidly, and have once again stabilized at about a 55% chance of restoration by July 1, 30% by June 26 and 12% by June 19. That seems modestly higher than I would put those numbers, but not unreasonable. Every day that Fable remains unavailable further damages America, its cyber defenses, its productivity and the world's trust in its AI and supposed ‘tech stack.’ Every day that Mythos remains unavailable is a day the free world's top companies and cyber defenders lose in their race against the avalanche headed their way. Mostly we have learned and confirmed more about exactly what happened. We know more about what Amazon did, what the official letter said, what the supposed ‘jailbreak’ was (literally, and I am not making this up, ‘fix this code’) and more. It is all about as stupid as it could have been. Table of [...] --- Outline: (01:22) There Was No Fable Jailbreak (07:16) If This Jailbreak Was Real It Would Be Trivial To Prove It (08:35) No Eyes (09:41) What The Letter Actually Said (11:29) Anthropic Cannot Challenge This But If It Did Then It Plausibly Wins (13:28) What Happened At Amazon (17:43) This Was Not About Chinese Access (18:01) Absolute Discretion And Ad Hockery Is Not Deregulation (20:43) All Of American AI Is Permanently Damaged As This Continues (22:14) Dean Ball Gives His Interpretation (25:03) Again, Yes, I Do Think Anthropic Should Have Taken Fable Down (28:02) To What Extent Was This A Deliberate Attack? (32:55) The Next Chapter For Fable (36:59) Our Continuing Coverage --- First published: June 17th, 2026 Source: https://www.lesswrong.com/posts/HaHzwvhbWam4n8hJB/the-once-and-future-fable-3-fix-this-code [https://www.lesswrong.com/posts/HaHzwvhbWam4n8hJB/the-once-and-future-fable-3-fix-this-code?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration] --- Narrated by TYPE III AUDIO [https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=lesswrong&utm_campaign=ai_narration]. --- Images from the article: Infographic showing export controls on cybersecurity capabilities, emphasizing defensive strategy over offensive restrictions. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/rpn4dotj5emydy6c22za]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/rpn4dotj5emydy6c22za ---------------------------------------- Woman wearing black t-shirt with pink text reading [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/hwsx3kharf5j7gvhi150]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/hwsx3kharf5j7gvhi150 ---------------------------------------- Person speaking with text overlay reading [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/zytxutgiwf150p0hudjl]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/zytxutgiwf150p0hudjl ---------------------------------------- Meme with three panels showing men discussing confessing versus bragging. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/gr3qri4e6mbs3rrpscih]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/gr3qri4e6mbs3rrpscih ---------------------------------------- A user with a [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/dzubbbeitrms7wdaec1c]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/dzubbbeitrms7wdaec1c ---------------------------------------- Distinguished man with white hair in dark three-piece suit seated at table. [https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/lrz4qzl2hjfrmey54k0z]https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/HaHzwvhbWam4n8hJB/lrz4qzl2hjfrmey54k0z Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts [https://pocketcasts.com/], or another podcast app.

17 jun 202637 min