Our 189th episode with a summary and discussion of last week's big AI news!

Hosted by Andrey Kurenkov and Jeremie Harris.
Feel free to email us your questions and feedback at [email protected] and/or [email protected]

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.

In this episode:

* OpenAI's acquisition of chat.com and internal shifts, including hardware lead hire and hardware model leaks, signal significant strategy pivots and challenges with model scaling and security. 
* Saudi Arabia plans a $100 billion AI initiative aiming to rival UAE's tech hub, highlighting the region's escalating AI investments. 
* U.S. penalties on GlobalFoundries for violating sanctions against SMIC underline ongoing challenges in enforcing AI-chip export controls. 
* Anthropic collaborates with Palantir and AWS to integrate CLAWD into defense environments, marking a significant policy shift for the company.

Sponsors:

• The Generator - An interdisciplinary AI lab empowering innovators from all fields to bring visionary ideas to life by harnessing the capabilities of artificial intelligence.
• The AI safety book “Uncontrollable" which is not a doomer book, but instead lays out the reasonable case for AI safety and what we can do about it. Max TEGMARK said that “Uncontrollable” is a captivating, balanced, and remarkably up-to-date book on the most important issue of our time" - find it on Amazon today!

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:01:28) News Preview
• (00:02:10) Response to listener comments
• (00:05:02) Sponsor Break
• Tools & Apps
  • (00:07:31) OpenAI Introduces ‘Predicted Outputs’ Feature: Speeding Up GPT-4o by ~5x for Tasks like Editing Docs or Refactoring Code
  • (00:11:55) Anthropic’s Haiku 3.5 surprises experts with an “intelligence” price increase
  • (00:17:10) Introducing FLUX1.1 [pro] Ultra and Raw Modes
  • (00:19:11) X is testing a free version of Grok AI chatbot in select regions
• Applications & Business
  • (00:21:39) OpenAI acquired Chat.com
  • (00:23:40) Saudis Plan $100 Billion AI Powerhouse to Rival UAE Tech Hub
  • (00:28:28) Meta’s former hardware lead for Orion is joining OpenAI
  • (00:31:38) OpenAI Accidentally Leaked Its Upcoming o1 Model to Anyone With a Certain Web Address
  • (00:35:50) Nvidia Rides AI Wave to Pass Apple as World’s Largest Company
• Projects & Open Source
  • (00:37:53) ‘Unrestricted’ AI group Nous Research launches first chatbot — with guardrails
  • (00:41:48) FrontierMath: The Benchmark that Highlights AI’s Limits in Mathematics
  • (00:46:29) Hunyuan-Large: An Open-Source MoE Model with 52 Billion Activated Parameters by Tencent
• Research & Advancements
  • (00:49:55) Applying “Golden Gate Claude” mechanistic interpretability techniques to protein language models.
  • (00:58:3) Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA
  • (01:05:55) From Naptime to Big Sleep: Using Large Language Models To Catch Vulnerabilities In Real-World Code
  • (01:10:22) OpenAI reportedly developing new strategies to deal with AI improvement slowdown
• Policy & Safety
  • (01:19:52)  What Donald Trump’s Win Means For AI
  • (01:28:44) Fab Whack-A-Mole: Chinese Companies are Evading U.S. Sanctions
  • (01:33:57) US fines GlobalFoundries for shipping chips to sanctioned Chinese firm
  • (01:36:55) Anthropic teams up with Palantir and AWS to sell its AI to defense customers
• (01:39:23) Outro

Our 188th episode with a summary and discussion of last week's big AI news!

Hosted by Andrey Kurenkov and Jeremie Harris.
Feel free to email us your questions and feedback at [email protected] and/or [email protected]

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.

This episode was sponsored by The Generator.
If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

In this episode:
* Meta's open-source models utilized by China's military prompt regulatory adjustments; US agencies gain access to counterbalance. 
* OpenAI partners with Broadcom and AMD to develop custom AI hardware, aiming for profitability and reducing inference costs. 
* Physical Intelligence unveils a generalist robot control policy with a $400M funding boost, showcasing significant advancements in zero-shot task performance. 
* New U.S. regulation mandates quarterly reporting for large AI model training and computing cluster acquisitions, aiming to bolster national security.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:02:16) News Preview
• (00:03:05) Response to listener comments / corrections
• (00:05:00) Sponsor Break
• Tools & Apps
  • (00:06:28) OpenAI’s search engine is now live in ChatGPT
  • (00:12:18) Image Playground, ChatGPT, and more Apple Intelligence features roll out in beta
  • (00:14:34) GitHub Copilot will support models from Anthropic, Google, and OpenAI
  • (00:19:00) Introducing the analysis tool in Claude.ai
  • (00:21:34) ElevenLabs Introduces Voice Design: A New AI Feature that Generates a Unique Voice from a Text Prompt Alone
  • (00:24:18) Midjourney's new web editor lets you tweak images uploaded from your PC
  • (00:26:02) Watch out, Midjourney — Recraft just announced new AI image generator model
• Applications & Business
  • (00:29:57) Meta strikes multi-year AI deal with Reuters
  • (00:33:15) OpenAI will start using AMD chips and could make its own AI hardware in 2026
  • (00:40:47) Elon Musk's xAI in talks to raise funding valuing it at $40 billion, WSJ reports
  • (00:46:07) Physical Intelligence, a Robot A.I. Specialist, Raises Millions From Bezos
  • (00:48:32) Waymo ramps up robotaxi push with $5.6 bn in funding
  • (00:49:11) Alphabet's Waymo Serving Over 150,000 Paid Robotaxi Rides Every Week Now, Surging 50% In 2 Months
• Projects & Open Source
  • (00:51:23) Meta AI Silently Releases NotebookLlama: An Open Version of Google’s NotebookLM
  • (00:54:59) Meta Releases Quantized Llama 3.2 with 4x Inference Speed on Android Phones
  • (00:59:16) OpenAI Releases SimpleQA: A New AI Benchmark that Measures the Factuality of Language Models
• Research & Advancements
  • (01:08:19) This Is a Glimpse of the Future of AI Robot
  • (01:15:06) Can Language Models Replace Programmers? REPOCOD Says 'Not Yet'
  • (01:19:01) Brain-like Functional Organization within Large Language Models
  • (01:21:20) Decart’s AI simulates a real-time, playable version of Minecraft
  • (01:25:39) Raising the bar on SWE-bench Verified with Claude 3.5 Sonnet
• Policy & Safety
  • (01:29:06) Commerce just proposed the most significant federal AI regulation to date – and no one noticed
  • (01:35:04)Anthropic warns of AI catastrophe if governments don't regulate in 18 months
  • (01:39:32) Open Source Bites Back as China’s Military Makes Full Use of Meta AI
  • (01:46:35) Meta says it’s making its Llama models available for US national security applications
• (01:48:16) Outro

Our 187th episode with a summary and discussion of last week's big AI news, now with Jeremie co-hosting once again!

Hosted by Andrey Kurenkov and Jeremie Harris. Feel free to email us your questions and feedback at [email protected] and/or [email protected]

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.

This episode was sponsored by The Generator.

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:03:07) Response to listener comments / corrections
• (00:05:13) Sponsor Read)
• Tools & Apps
  • (00:06:22) Anthropic’s latest AI update can use a computer on its own
  • (00:18:09) AI video startup Genmo launches Mochi 1, an open source rival to Runway, Kling, and others
  • (00:20:37) Canva has a shiny new text-to-image generator
  • (00:23:35) Canvas Beta brings Remix, Extend, and Magic Fill to Ideogram users
  • (00:26:16) StabilityAI releases Stable Diffusion 3.5 
  • (00:28:27) Bringing Agentic Workflows into Inflection for Enterprise
• Applications & Business
  • (00:32:35) Crusoe’s $3.4B joint venture to build AI data center campus with up to 100,000 GPUs
  • (00:39:08) Anthropic reportedly in early talks to raise new funding on up to $40B valuation
  • (00:45:47) Longtime policy researcher Miles Brundage leaves OpenAI
  • (00:49:53) NVIDIA’s Blackwell GB200 AI Servers Ready For Mass Deployment In December
  • (00:52:41) Foxconn building Nvidia superchip facility in Mexico, executives say
  • (00:55:27) xAI, Elon Musk’s AI startup, launches an API
• Projects & Open Source
  • (00:58:32) INTELLECT-1: The First Decentralized 10-Billion-Parameter AI Model Training
  • (01:06:34) Meta FAIR Releases Eight New AI Research Artifacts—Models, Datasets, and Tools to Inspire the AI Community
  • (01:10:02) Google DeepMind is making its AI text watermark open source
• Research & Advancements
  • (01:13:21) OpenAI researchers develop new model that speeds up media generation by 50X
  • (01:17:54) How much AI compute is out there, and who owns it?
  • (01:25:28) Rewarding Progress: Scaling Automated Process Verifiers for LLM Reasoning
  • (01:33:30) Inference Scaling for Long-Context Retrieval Augmented Generation
• Policy & Safety
  • (01:41:50) Announcing our updated Responsible Scaling Policy
  • (01:48:52) Anthropic is testing AI’s capacity for sabotage
  • (01:56:30) OpenAI asked US to approve energy-guzzling 5GW data centers, report says
  • (02:00:05) US Probes TSMC’s Dealings with Huawei
  • (02:03:03) TikTok owner ByteDance taps TSMC to make its own AI GPUs to stop relying on Nvidia — the company has reportedly spent over $2 billion on Nvidia AI GPUs
• (02:06:37) Outro

Our 186th episode with a summary and discussion of last week's big AI news! With hosts Andrey Kurenkov and guest host Jon Krohn from the SuperDataScience Podcast.

Check out Jon’s upcoming agent-focused event here - AI Catalyst: Agentic Artificial Intelligence

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:04:14) News Preview
• (00:05:28) Response to listener comments / corrections
• Tools & Apps
  • (00:07:10) Adobe’s AI video model is here, and it’s already inside Premiere Pro
  • (00:11:52) Adobe teases AI tools that build 3D scenes, animate text, and make distractions disappear
  • (00:15:43) Adobe’s Project Super Sonic uses AI to generate sound effects for your videos
  • (00:17:05) YouTube expands AI audio generation tool to all U.S. creators
  • (00:20:29) All Gemini users can now generate images with Imagen 3
  • (00:22:27) Meta AI will launch in six more countries today, including the UK
  • (00:24:27) OpenAI Unveils Secret Meta Prompt—And It’s Very Different From Anthropic's Approach
• Applications & Business
  • (00:27:46) Tesla’s big ‘We, Robot’ event criticized for ‘parlor tricks’ and vague timelines for robots, Cybercab, Robovan
  • (00:37:25) OpenAI announces content deal with Hearst, including content from Cosmopolitan, Esquire and the San Francisco Chronicle
• Projects & Open Source
  • (00:47:59) OpenR: An Open-Source AI Framework Enhancing Reasoning in Large Language Models
  • (00:49:54) MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering
  • (00:56:29) OpenAI Releases Swarm: An Experimental AI Framework for Building, Orchestrating, and Deploying Multi-Agent Systems
• Research & Advancements
  • (00:59:23) Nobel Physics Prize Awarded for Pioneering A.I. Research by 2 Scientists
  • (01:05:22) Nobel Prize in Chemistry Goes to 3 Scientists for Predicting and Creating Proteins
  • (01:09:09) LLMs can’t perform “genuine logical reasoning,” Apple researchers suggest
  • (01:13:05) GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models
• Policy & Safety
  • (01:14:34) Anthropic CEO goes full techno-optimist in 15,000-word paean to AI
  • (01:23:04) Google will help build seven nuclear reactors to power its AI systems
  • (01:24:11) LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations
• Synthetic Media & Art
  • (01:26:26) Adobe Pushes Content Authenticity Forward With a Free Web App Designed for Creators
• (01:29:13) Outro

Our 185th episode with a summary and discussion of last week's big AI news! With hosts Andrey Kurenkov and guest host Gavin Purcell from the AI for Humans podcast.

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

In this episode:

• Meta's MovieGen introduces innovative features in AI video generation, alongside OpenAI's real-time speech API and expanded ChatGPT capabilities.
• Mio's foundation model and Apple's Depth Pro enhance multimodal AI inputs and precise 3D imaging for AR, VR, and robotics.
• Microsoft and OpenAI's strategic advancements highlight significant financial moves and AI enhancements, including Microsoft's enhanced Copilot.
• AI policy discussions intensify as California's vetoed bill sparks debates on regulation, alongside Google's $1 billion investment to expand AI infrastructure in Thailand.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:02:51) Response to listener comments / corrections
• Tools & Apps
  • (00:03:48) Meta announces Movie Gen, an AI-powered video generator
  • (00:14:28) OpenAI launches new ‘Canvas’ ChatGPT interface tailored to writing and coding projects
  • (00:19:31) OpenAI’s DevDay brings Realtime API and other treats for AI app developers
  • (00:24:43) Black Forest Labs releases Flux 1.1 Pro and an API
  • (00:28:30) Microsoft gives Copilot a voice and vision in its biggest redesign yet
  • (00:32:36) Pika 1.5 is now live — AI video generator just got major upgrades
• Applications & Business
  • (00:37:49) OpenAI closes the largest VC round of all time
  • (00:45:23) Google brings ads to AI Overviews as it expands AI’s role in search
  • (00:51:05) Anthropic hires OpenAI co-founder Durk Kingma
  • (00:51:49) OpenAI’s newest creation is raising shock, alarm, and horror among staffers: a new logo
  • (00:53:45) Waymo to add Hyundai EVs to robotaxi fleet under new multiyear deal
  • (00:57:28) Cerebras, an A.I. Chipmaker Trying to Take On Nvidia, Files for an I.P.O.
  • (00:59:18) Y Combinator is being criticized after it backed an AI startup that admits it basically cloned another AI startup
• Research & Advancements
  • (01:03:30) Were RNNs All We Needed?
  • (01:06:52) MIO: A Foundation Model on Multimodal Tokens
  • (01:09:20) Apple releases Depth Pro, an AI model that rewrites the rules of 3D vision
• Policy & Safety
  • (01:13:08) California Governor Vetoes Sweeping A.I. Legislation
  • (01:18:02) Judge blocks California’s new AI law in case over Kamala Harris deepfake Musk reposted
  • (01:20:41) Google to invest $1 billion in Thailand to build a data center and accelerate AI growth
• Synthetic Media & Art
  • (01:21:58) AI reading coach startup Ello now lets kids create their own stories
• (01:25:13) Outro

Our 184th episode with a summary and discussion of last week's big AI news! With hosts Andrey Kurenkov and guest host Jon Krohn.

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

In this episode:

• OpenAI, Meta, and Google are enhancing their AI assistants with advanced voice modes, while Meta released Llama 3.2, an open-source model capable of processing both images and text.
• Significant AI infrastructure developments include Grok's partnership with Aramco for a massive data center in Saudi Arabia, and Microsoft's plan to power data centers using a reopened Three Mile Island nuclear plant.
• Recent research shows chain-of-thought prompting is most effective for math and symbolic reasoning, while OpenAI's GPT-4 with vision capabilities is being integrated into Perplexity AI's search platform.
• AI is being rapidly integrated into various sectors, with examples including ChartWatch reducing unexpected hospital deaths, Snapchat and YouTube introducing AI video generation tools, and Lionsgate partnering with Runway for AI-assisted film production.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:04:45) Response to listener comments / corrections
• Tools & Apps
  • (00:07:46) OpenAI rolls out Advanced Voice Mode with more voices and a new look
  • (00:13:32) Meta’s AI can now talk to you in the voices of Awkwafina, John Cena, and Judi Dench
  • (00:17:11) Gemini’s chatty voice mode is out now for free on Android
  • (00:21:30) AI video rivalry intensifies as Luma announces Dream Machine API hours after Runway
  • (00:23:35) Copilot Wave 2 supercharges productivity with AI across all your Microsoft 365 apps
  • (00:25:56) Perplexity introduces new 'Reasoning' focus powered by OpenAI's o1
• Applications & Business
  • (00:33:47) OpenAI Execs Mass Quit as Company Removes Control From Non-Profit Board and Hands It to Sam Altman
  • (00:41:46) Sam Altman departs OpenAI’s safety committee
  • (00:43:04) Chip Startup Groq Backs Saudi AI Ambitions With Aramco Deal
  • (00:46:29) Grok’s image generator, Black Forest Labs, is raising $100M at a $1B valuation, say sources
  • (00:48:05) Pudu unveils super semi-humanoid robot with 8-hour battery, 10kg lift power 
  • (00:50:56) Amazon introduces Amelia, an AI assistant for third-party sellers
• Projects & Open Source
  • (00:52:58) Meta Releases Llama 3.2—and Gives Its AI a Voice
  • (00:56:52) Alibaba Unveils Ovis 1.6 – A New Multimodal Language Model
• Research & Advancements
  • (01:00:35) To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning
  • (01:06:14) LLMs Still Can't Plan; Can LRMs? A Preliminary Evaluation of OpenAI's o1 on PlanBench
  • (01:10:15) Norwegian Startup 1X Unveils AI World Model for Robot Training
  • (01:12:44) AI tool cuts unexpected deaths in hospital by 26%, Canadian study finds
• Policy & Safety
  • (01:15:47) Three Mile Island nuclear plant will reopen to power Microsoft data centers
  • (01:18:23) Governor Newsom signs bills to combat deepfake election content
  • (01:20:12) Governor Newsom signs bills to protect digital likeness of performers
  • (01:22:20) Startup behind “world’s first robot lawyer” to pay $193K for false ads, FTC says
• Synthetic Media & Art
  • (01:24:49) Snap is introducing an AI video generation tool for creators
  • (01:25:57) YouTube Shorts to integrate Veo, Google’s AI video model 
  • (01:26:56) Lionsgate Signs Deal With AI Company Runway, Hopes That AI Can Eliminate Storyboard Artists and VFX Crews

(01:28:01) Outro

Our 183rd episode with a summary and discussion of last week's big AI news! With hosts Andrey Kurenkov and Jeremie Harris.

Note: once again, apologies from Andrey on this one coming out late. Starting with the next one we should be back to a regular(ish) release schedule.

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/. If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

In this episode:

- OpenAI's O1 and O1 mini models boast advanced reasoning and longer responses. 

- Adobe adds video generation to Firefly, Anthropic launches AI safety-focused Claude enterprise.

- LLAMA3 8B excels with synthetic tokens, AI-generated ideas deemed more novel.

- New AI forecasting bot competes with veteran human forecasters.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:02:38) Response to listener comments / corrections
• Tools & Apps
  • (00:04:00) OpenAI releases o1, its first model with ‘reasoning’ abilities
  • (00:22:06) Adobe says video generation is coming to Firefly this year
  • (00:25:27) Anthropic launches Claude Enterprise with more security, admin controls
  • (00:27:55) Tell Replit's AI Agent Your App Idea, and It'll Code and Deploy It for You
• Applications & Business
  • (00:31:42) OpenAI Fundraising Set to Vault Startup’s Valuation to $150 Billion
  • (00:37:03) OpenAI Hits 1 Million Paid Users For Business Versions of ChatGPT
  • (00:39:59) TSMC Arizona achieves production yields similar to those at its fabs in Taiwan, says report
  • (00:45:14) Japan’s Sakana AI partners Nvidia for research, raises $100M
• Projects & Open Source
  • (00:47:31) The fable of Reflection 70B
  • (00:53:08) DataGemma: Using real-world data to address AI hallucinations
  • (00:56:31) DeepSeek-V2.5 wins praise as the new, true open source AI model leader
• Research & Advancements
  • (01:01:35) Google DeepMind Launches AlphaProteo , an AI Model for Generating Proteins 
  • (01:08:27) Synthetic continued pretraining
  • (01:16:55) A Large-Scale Human Study with 100+ NLP Researchers
  • (01:23:55) Superhuman Automated Forecasting
• Policy & Safety
  • (01:28:05) OpenAI o1 System Card
  • (01:36:58) US roll outs new export controls aimed at restricting China’s chip industry
  • (01:39:50) Employee Call for Governor Newsom to Sign SB 1047
  • (01:42:34) Taylor Swift says AI version of herself falsely endorsing Trump 'conjured up my fears'
• Synthetic Media & Art
  • (01:44:42) YouTube is developing AI detection tools for music and faces, plus creator controls for AI training
• (01:45:08) Outro

Our 182nd episode with a summary and discussion of last week's big AI news! With hosts Andrey Kurenkov and Jeremie Harris.

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/. If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

Sponsors:

- Agent.ai is the global marketplace and network for AI builders and fans. Hire AI agents to run routine tasks, discover new insights, and drive better results. Don't just keep up with the competition—outsmart them. And leave the boring stuff to the robots 🤖

- Pioneers of AI, is your trusted guide to this emerging technology. Host Rana el Kaliouby (RAH-nuh el Kahl-yoo-bee) is an AI scientist, entrepreneur, author and investor exploring all the opportunities and questions AI brings into our lives. Listen to Pioneers of AI, with new episodes every Wednesday, wherever you tune in.

In this episode:

- OpenAI's move into hardware production and Amazon's strategic acquisition in AI robotics.
- Advances in training language models with long-context capabilities and California's pending AI regulation bill.
- Strategies for safeguarding open weight LLMs against adversarial attacks and China's rise in chip manufacturing.
- Sam Altman's infrastructure investment plan and debates on AI-generated art by Ted Chiang.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:05:15) Response to listener comments / corrections

• Tools & Apps
  • (00:07:32) Amazon Picks Anthropic to Power Alexa 2.0
  • (00:12:12) Forget Sora — MiniMax is a new realistic AI video generator and it’s seriously impressive
• Applications & Business
  • (00:14:56) Ilya Sutskever’s startup, Safe Superintelligence, raises $1B
  • (00:22:20) TSMC’s A16 Process Creates a Buzz Before Mass Production, as OpenAI Reportedly Secures Capacity
  • (00:29:13) Amazon hires the founders of AI robotics startup Covariant
  • (00:33:33) OpenAI weighs changes to corporate structure amid latest funding talks
  • (00:37:43) Chinese GPU-maker XCT, once valued at $2.1B, is on the verge of collapse — shareholders now suing founder
  • (00:40:34) TSMC aims to ready next-gen silicon photonics for AI in 5 years
• Projects & Open Source
  • (00:44:10) Alibaba releases new AI model Qwen2-VL that can analyze videos more than 20 minutes long
• Research & Advancements
  • (00:48:32) Fire-Flyer AI-HPC: A Cost-Effective Software-Hardware Co-Design for Deep Learning
  • (00:55:52) 100M Token Context Windows
  • (01:03:50) Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Sampling
  • (01:06:16) AnyGraph : An Effective and Efficient Graph Foundation Model Designed to Address the Multifaceted Challenges of Structure and Feature Heterogeneity Across Diverse Graph Datasets
• Policy & Safety
  • (01:08:16) California Legislature Approves Bill Proposing Sweeping A.I. Restrictions
  • (01:11:14) Tamper-Resistant Safeguards for Open-Weight LLMs
  • (01:17:12) China's chip capabilities just 3 years behind TSMC, teardown shows
  • (01:20:50) China Threatens to Cut Off ASML Over New US Chip Curbs
  • (01:23:22) Altman Infrastructure Plan Aims to Spend Tens of Billions in US
• Synthetic Media & Art
  • (01:26:23) NaNoWriMo is in disarray after organizers defend AI writing tools
  • (01:28:54) Tom Hanks warns followers to be wary of 'fraudulent' ads using his likeness through AI
  • (01:30:48) Why A.I. Isn’t Going to Make Art
• (01:34:28) Outro

Our 181st episode with a summary and discussion of last week's big AI news!

With hosts Andrey Kurenkov and Jeremie Harris

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

In this episode:

- Google's AI advancements with Gemini 1.5 models and AI-generated avatars, along with Samsung's lithography progress. 
- Microsoft's Inflection usage caps for Pi, new AI inference services by Cerebrus Systems competing with Nvidia. 
- Biases in AI, prompt leak attacks, and transparency in models and distributed training optimizations, including the 'distro' optimizer. 
- AI regulation discussions including California’s SB1047, China's AI safety stance, and new export restrictions impacting Nvidia’s AI chips.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:03:08)Response to listener comments / corrections
• Tools & Apps
  • (00:09:19) Google’s custom AI chatbots have arrived
  • (00:12:52) Google releases three new experimental AI models
  • (00:17:14) Google Gemini will let you create AI-generated people again
  • (00:22:32) Five months after Microsoft hired its founders, Inflection adds usage caps to Pi
  • (00:26:42:) Plaud takes a crack at a simpler AI pin
• Applications & Business
  • (00:30:31) Cerebras Systems throws down gauntlet to Nvidia with launch of ‘world’s fastest’ AI inference service
  • (00:41:06) Nvidia announces $50 billion stock buyback
  • (00:46:24) OpenAI in talks to raise funding that would value it at more than $100 billion
  • (00:50:44) OpenAI Aims to Release New AI Model, ‘Strawberry,’ in Fall
  • (00:52:53) 3 Co-Founders Leave French AI Startup H Amid ‘Operational Differences’
  • (00:57:29) Samsung to Adopt High-NA Lithography Alongside Intel, Ahead of TSMC
  • (01:02:11) Unitree's $16,000 G1 could become the first mainstream humanoid robot
• Projects & Open Source
  • (01:04:59) Meta leads open-source AI boom, Llama downloads surge 10x year-over-year
  • (01:09:08) A_Preliminary_Report_on_DisTrO.
• Research & Advancements
  • (01:13:56) Diffusion Models Are Real-Time Game Engines
  • (01:23:18) LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yet
  • (01:32:21) Interviewing AI researchers on automation of AI R&D
  • (01:40:33) Anthropic releases AI model system prompts, winning praise for transparency
• Policy & Safety
  • (01:47:12) U.S. AI Safety Institute Signs Agreements Regarding AI Safety Research, Testing and Evaluation With Anthropic and OpenAI
  • (01:50:46) China’s Views on AI Safety Are Changing—Quickly
  • (01:56:27) Poll: 7 in 10 Californians Support SB1047, Will Blame Governor Newsom for AI-Enabled Catastrophe if He Vetoes
  • (02:01:31) Elon Musk voices support for California bill requiring safety tests on AI models
  • (02:03:55) Chinese Engineers Reportedly Accessing NVIDIA’s High-End AI Chips Through Decentralized “GPU Rental Services”
  • (02:08:25) U.S. gov't tightens China restrictions on supercomputer component sales
• Synthetic Media & Art
  • (02:11:13) Actors Say AI Voice-Over Generator ElevenLabs Cloned Likenesses
• (02:14:06) Outro

Our 180th episode with a summary and discussion of last week's big AI news!

With hosts Andrey Kurenkov (https://twitter.com/andrey_kurenkov) and Jeremie Harris (https://twitter.com/jeremiecharris)

If you would like to get a sneak peek and help test Andrey's generative AI application, go to Astrocade.com to join the waitlist and the discord.

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

Episode Highlights:

• Ideogram AI's new features, Google's Imagine 3, Dream Machine 1.5, and Runway's Gen3 Alpha Turbo model advancements.
• Perplexity's integration of Flux image generation models and code interpreter updates for enhanced search results. 
• Exploration of the feasibility and investment needed for scaling advanced AI models like GPT-4 and Agent Q architecture enhancements.
• Analysis of California's AI regulation bill SB1047 and legal issues related to synthetic media, copyright, and online personhood credentials.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:01:08) Response to Listener Comments / Corrections
• Tools & Apps
  • (00:03:58) Ideogram AI expands its features with v2 model and color palette options
  • (00:07:48) Google Releases Powerful AI Image Generator You Can Use for Free
  • (00:11:41) Perplexity adds Flux.1 model for Pro users alongside Playground v3 update
  • (00:13:58) Luma drops Dream Machine 1.5 — here’s what’s new
  • (00:17:49) Runway’s Gen-3 Alpha Turbo is here and can make AI videos faster than you can type
  • (00:20:21) Perplexity’s latest update improves code interpreter, charts included
• Applications & Business
  • (00:24:14) AMD buying server maker ZT Systems for $4.9 billion as chipmakers strengthen AI capabilities
  • (00:28:55) Ars Technica content is now available in OpenAI services
  • (00:34:08) Anysphere, a GitHub Copilot rival, has raised $60M Series A at  $400M valuation from a16z, Thrive, sources say
  • 00:38:32 Stability AI appoints new Chief Technology Officer
  • (00:41:45) Cruise’s robotaxis are coming to the Uber app in 2025
• Projects & Open Source
  • (00:44:16) AI21 Introduces the Jamba Model Family: The most powerful and efficient long-context models for the enterprise
  • (00:53:47) Microsoft reveals Phi-3.5 — this new small AI model outperforms Gemini and GPT-4o
  • (00:57:33) Nvidia’s Llama-3.1-Minitron 4B is a small language model that punches above its weight
  • (01:00:58) Open source Dracarys models ignite generative AI fired coding
• Research & Advancements
  • (01:12:35) Can AI Scaling Continue Through 2030?
  • (01:15:35) Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents
  • (01:23:58) Transformers to SSMs: Distilling Quadratic Knowledge to Subquadratic Models
  • (01:31:18) Loss of plasticity in deep continual learning
• Policy & Safety
  • (01:38:20) California weakens bill to prevent AI disasters before final vote, taking advice from Anthropic
  • (01:48:14) Personhood credentials: Artificial intelligence and the value of privacy-preserving tools to distinguish who is real online
  • (01:52:44) Showing SAE Latents Are Not Atomic Using Meta-SAEs
• Synthetic Media & Art
  • (01:58:33) Authors sue Claude AI chatbot creator Anthropic for copyright infringement
  • (01:59:32) Artists’ lawsuit against Stability AI and Midjourney gets more punch
• (02:01:43) Outro

Our 179th episode with a summary and discussion of last week's big AI news!

With hosts Andrey Kurenkov (https://twitter.com/andrey_kurenkov) and Jeremie Harris (https://twitter.com/jeremiecharris)

If you would like to get a sneak peek and help test Andrey's generative AI application, go to Astrocade.com to join the waitlist and the discord.

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/

If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.

Email us your questions and feedback at [email protected] and/or [email protected]

Episode Highlights:

- Grok 2's beta release features new image generation using Black Forest Labs' tech.

- Google introduces Gemini Voice Chat Mode available to subscribers and integrates it into Pixel Buds Pro 2.

- Huawei's Ascend 910C AI chip aims to rival NVIDIA's H100 amidst US export controls.

- Overview of potential risks of unaligned AI models and skepticism around SingularityNet's AGI supercomputer claims.

Timestamps + Links:

• (00:00:00) Intro / Banter
• (00:02:15) Response to listener comments / corrections
• Tools & Apps
  • (00:04:24) Grok-2 is out in beta, now with added AI image generation
  • (00:11:28) OpenAI reveals an updated GPT-4o model - but can't quite explain how it's better
  • (00:13:48) Google Gemini’s voice chat mode is here
  • (00:16:18) Google’s Pixel Buds Pro 2 bring Gemini to your ears
  • (00:19:55) Google’s AI-generated search summaries change how they show their sources
  • (00:23:13) Prompt Caching is Now Available on the Anthropic API for Specific Claude Models
• Applications & Business
  • (00:26:56) Meet Black Forest Labs, the startup powering Elon Musk’s unhinged AI image generator
  • (00:26:56) Huawei readies new AI chip to challenge Nvidia in China, WSJ reports
  • (00:37:53) ASML and Imec Announce High-NA Lithography Breakthrough
  • (00:43:07) Chinese startup WeRide gets nod to test robotaxis with passengers in California
  • (00:45:49) Perplexity’s popularity surges as AI search start-up takes on Google
  • (00:51:55) Lisa Su formally welcomes Silo AI team to AMD after completing $665 million acquisition
• Projects & Open Source
  • (00:54:31) FalconMamba 7B Released: The World’s First Attention-Free AI Model with 5500GT Training Data and 7 Billion Parameters
  • (00:59:25) OpenAI has introduced SWE-bench Verified to evaluate AI performance
  • (01:04:21) Nous Research presents Hermes 3
    
  • (01:11:07) New supercomputing network could lead to AGI, scientists hope, with 1st node coming online within weeks
• Research & Advancements
  • (01:14:40) The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
  • (01:30:24) Imagen 3
  • (01:32:48) The Data Addition Dilemma
  • (01:37:35) LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs
• Policy & Safety
  • (01:40:55) MIT researchers release a repository of AI risks
  • (01:44:14) Elon Musk addresses power issues at xAI supercomputer facility in Memphis
  • (01:46:52) FCC Proposes New Rules on AI-Powered Robocalls
  • Governments Adjust Policies Amid Flood of AI Record Requests
• Synthetic Media & Art
  • (01:48:21) SAG-AFTRA Strikes Groundbreaking AI Digital Voice Replica Pact With Startup Firm Narrativ
  • (01:51:52) How ‘Deepfake Elon Musk’ Became the Internet’s Biggest Scammer
• (01:56:21) AI Song Outro