#217 - ChatGPT Agent, Kimi k2, Hiring Drama
1Digest
This podcast episode, hosted by Andrey Krankov and John Crone, covers significant developments in the AI landscape. It begins by introducing the hosts and John's new AI consulting firm, Y Carat. The episode then delves into new AI tools, focusing on OpenAI's ChatGPT agent (its capabilities and safety concerns), the open-source Kimi K2 model, Amazon's Kuro AI tool, and Anthropic's Claude. The discussion shifts to the competitive AI business environment, highlighting acquisitions like Cognition's purchase of Windsurf (following a cancelled OpenAI deal) and the intense hiring battles between tech giants like Meta, Google, and Anthropic. The episode also features insights from a former OpenAI engineer's blog post about the company's culture and rapid growth. Research challenges are addressed, specifically concerning reinforcement learning and data contamination issues in models like Quen. Finally, the podcast explores policy and safety concerns, including large government AI contracts, the push for AI regulation, ethical dilemmas surrounding AI-generated explicit content, and the use of facial recognition technology by ICE. The resolution of the video game actors' strike is also mentioned.
Outlines
Introduction and AI Landscape Overview
Introduces the hosts and their backgrounds in AI, mentions John's new consulting firm (Y Carat), and previews the episode's focus on new AI tools, business developments (acquisitions and hiring), research, and policy/safety concerns.
New AI Tools and Open Source Advancements
Details OpenAI's ChatGPT agent, its capabilities, and safety concerns; discusses the Kimi K2 open-source model, Amazon's Kuro AI development tool, and Anthropic's Claude usage limits; mentions Mistral's enhanced research capabilities.
The AI Acquisition and Hiring Frenzy
Covers the cancelled OpenAI-Windsurf deal, Google's hiring of Windsurf's CEO, Cognition's acquisition of Windsurf, Anthropic's re-hiring, Meta's aggressive hiring, and Mira Murati's new company funding.
Inside OpenAI and Industry Culture
Explores insights from a former OpenAI engineer's blog post regarding the company's culture, rapid growth, and work environment.
Research Challenges and Data Contamination
Discusses a research paper highlighting issues with reinforcement learning and data contamination in AI models, particularly concerning the Quen model family and benchmark leakage.
AI Policy, Safety, and Ethical Concerns
Covers large AI contracts awarded by the Department of Defense, the push for AI regulation in California, ethical issues surrounding AI-generated explicit content, concerns about ICE's use of facial recognition technology, and the resolution of the video game actors' strike.
Keywords
ChatGPT Agent
OpenAI's new ChatGPT mode combining web browsing and task execution, exceeding previous models in performance but raising safety concerns.
Kimi K2
A trillion-parameter open-source AI model from Alibaba, rivaling proprietary models in coding benchmarks.
Agentic AI
AI systems capable of autonomous task execution, involving web browsing, code generation, and other complex actions.
LLM Benchmarks
Standardized tests for large language models, subject to data leakage and model-specific biases.
AI Hiring Frenzy
Intense competition among tech companies to acquire top AI talent, reflecting the rapid growth of the AI field.
AI Regulation
Growing calls for government regulation of AI to address ethical and safety concerns.
AI-Generated Content
AI-created content, raising ethical concerns regarding non-consensual explicit material and potential misuse.
Facial Recognition
Technology used for identification, raising privacy and ethical concerns, particularly in law enforcement applications.
Reinforcement Learning
A machine learning technique facing challenges related to data contamination and benchmark leakage.
OpenAI
Leading AI research company, discussed in relation to its culture, hiring practices, and technological advancements.
Q&A
What are some of the most significant advancements in AI tools discussed in this podcast?
OpenAI's ChatGPT agent, the open-source Kimi K2 model, and Amazon's Kuro AI tool represent key advancements in agentic AI capabilities.
What are the major ethical concerns surrounding the rapid advancement of AI discussed in this episode?
The proliferation of AI-generated non-consensual explicit content, the potential misuse of facial recognition technology by law enforcement, and the broader societal impact of job displacement due to automation are highlighted.
What are the key trends in the AI business landscape, particularly regarding hiring and acquisitions?
Major tech companies are aggressively acquiring top AI talent, reflecting intense competition and rapid growth in the sector. Government agencies are also investing heavily in AI development.
Show Notes
Our 217th episode with a summary and discussion of last week's big AI news!
Recorded on 07/17/2025
Hosted by Andrey Kurenkov and guest co-host Jon Krohn.
Feel free to email us your questions and feedback at contact@lastweekinai.com and/or hello@gladstone.ai
Check out Jon's workshop on Agentic AI Engineering, and find his consultancy here.
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/.
In this episode:
- OpenAI's new ChatGPT agent: The episode begins with a detailed discussion on OpenAI's latest ChatGPT agent, which can control entire computers and perform a wide range of tasks, showcasing powerful performance benchmarks and potential applications in business and research.
- Major business moves in the AI space: Significant shifts include Google's acquisition of Windsurf's top talent after OpenAI's deal fell through, Cognition's acquisition of Windsurf, and several notable hires by Meta from OpenAI and Apple, highlighting intense competition in the AI industry.
- AI's ethical and societal impacts: The hosts discuss serious concerns like the rise of non-consensual explicit AI-generated images, ICE's use of facial recognition for large databases, and regulations aimed at controlling AI's potential misuse.
- Video game actors strike ends: The episode concludes with news that SAG-AFTRA's year-long strike for video game voice actors has ended after reaching an agreement on AI rights and wage increases, reflecting the broader impact of AI on the job market.
Timestamps + Links:
Tools & Apps
- (00:03:29 ) OpenAI’s new ChatGPT Agent can control an entire computer and do tasks for you
- (00:07:11 ) Alibaba-backed Moonshot releases new Kimi AI model that beats ChatGPT, Claude in coding — and it costs less
- (00:09:36 ) Amazon targets vibe-coding chaos with new 'Kiro' AI software development tool – GeekWire
- (00:12:33 ) Anthropic tightens usage limits for Claude Code – without telling users
- (00:15:51 ) Mistral's Le Chat chatbot gets a productivity push with new ‘deep research' mode | TechCrunch
- (00:17:46 ) I spent 24 hours flirting with Elon Musk’s AI girlfriend
- (00:21:32 ) Uber is close to completing its quest to become the ultimate robotaxi app | The Verge
Applications & Business
- (00:24:02 ) OpenAI’s Windsurf deal is off — and Windsurf’s CEO is going to Google | The Verge
- (00:28:09 ) Cognition, maker of the AI coding agent Devin, acquires Windsurf | TechCrunch
- (00:28:46 ) Anthropic hired back two of its employees — just two weeks after they left for a competitor. | The Verge
- (00:28:46 ) Another High-Profile OpenAI Researcher Departs for Meta | WIRED
- (00:28:46 ) Meta Hires Two Key Apple (AAPL) AI Experts After Poaching Their Boss - Bloomberg
- (00:31:31 ) Mira Murati's Thinking Machines Lab is worth $12B in seed round | TechCrunch
- (00:33:20 ) Lovable becomes a unicorn with $200M Series A just 8 months after launch | TechCrunch
- (00:34:55 ) SpaceX commits $2 billion to xAI as Musk steps up AI ambitions: Report | World News - Business Standard
Research & Advancements
- (00:35:59 ) A former OpenAI engineer describes what it’s really like to work there | TechCrunch
- (00:38:23 ) Reasoning or Memorization? Unreliable Results of Reinforcement Learning Due to Data Contamination
Policy & Safety
- (00:42:14 ) Anthropic, Google, OpenAI, xAI granted up to $200 million from DoD
- (00:43:08 ) California State Senator Scott Wiener Pushes Bill to Regulate AI Companies - Bloomberg
- (00:43:58 ) AI 'Nudify' Websites Are Raking in Millions of Dollars | WIRED
- (00:45:55 ) Inside ICE’s Supercharged Facial Recognition App of 200 Million Images
Synthetic Media & Art
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.
Table of contents
United States
