Our 205th episode with a summary and discussion of last week's big AI news!
Recorded on 03/28/2025
Hosted by <a href='https://twitter.com/andrey_kurenkov'>Andrey Kurenkov</a> and <a href='https://twitter.com/jeremiecharris'>Jeremie Harris</a>. 
Feel free to email us your questions and feedback at <a href='mailto:contact@lastweekinai.com'>contact@lastweekinai.com </a>and/or <a href='mailto:hello@gladstone.ai'>hello@gladstone.ai</a>
Read out our text newsletter and comment on the podcast at <a href='https://lastweekin.ai/'>https://lastweekin.ai/</a>.
<a href='https://discord.gg/nTyezGSKwP'>Join our Discord here!</a> https://discord.gg/nTyezGSKwP
In this episode:
<ul>
<li>OpenAI's new image generation capabilities represent significant advancements in AI tools, showcasing impressive benchmarks and multimodal functionalities.</li>
<li>OpenAI is finalizing a historic $40 billion funding round led by SoftBank, and Sam Altman shifts focus to technical direction while COO Brad Lightcap takes on more operational responsibilities.,</li>
<li>Anthropic unveils groundbreaking interpretability research, introducing cross-layer tracers and showcasing deep insights into model reasoning through applications on Claude 3.5.</li>
<li>New challenging benchmarks such as ARC AGI 2 and complex Sudoku variations aim to push the boundaries of reasoning and problem-solving capabilities in AI models.</li>
</ul>
Timestamps + Links:
<ul>
<li style="font-weight:400;">(00:00:00) Intro / Banter</li>
<li style="font-weight:400;">(00:01:01) News Preview</li>
<li style="font-weight:400;">Tools & Apps</li>
</ul>
<ul>
<li style="font-weight:400;">(00:02:46) <a href='https://blog.google/technology/google-deepmind/gemini-model-thinking-updates-march-2025/#gemini-2-5-thinking'>Gemini 2.5: Our most intelligent AI model</a></li>
<li style="font-weight:400;">(00:08:41) <a href='https://www.theverge.com/openai/635118/chatgpt-sora-ai-image-generation-chatgpt'>OpenAI rolls out image generation powered by GPT-4o to ChatGPT</a></li>
<li style="font-weight:400;">(00:16:14) <a href='https://the-decoder.com/ideogram-presents-version-3-0-of-its-ai-image-generation-system/'>Ideogram presents version 3.0 of its AI image generation system</a></li>
<li style="font-weight:400;">(00:19:20) <a href='https://decrypt.co/311375/new-reve-image-generator-beats-ai-art-heavyweights-midjourney-and-flux-at-a-penny-per-image'>New Reve Image Generator Beats AI Art Heavyweights MidJourney and Flux at a Penny Per Image</a></li>
<li style="font-weight:400;">(00:21:56) <a href='https://analyticsindiamag.com/ai-news-updates/alibaba-releases-qwen2-5-omni-adds-voice-and-video-modes-to-qwen-chat/'>Alibaba Releases Qwen2.5 Omni, Adds Voice and Video Modes to Qwen Chat</a></li>
<li style="font-weight:400;">(00:23:58) <a href='https://longportapp.com/en/news/232780347'>The official version of Tencent's Hunyuan Deep Thinking Model T1 is here, with fast articulation, instant responses, and a decoding speed increase of 2 times</a></li>
</ul>
<ul>
<li style="font-weight:400;">Applications & Business</li>
</ul>
<ul>
<li style="font-weight:400;">(00:25:45) <a href='https://www.bloomberg.com/news/articles/2025-03-26/openai-close-to-finalizing-its-40-billion-softbank-led-funding'>OpenAI Close to Finalizing $40 Billion SoftBank-Led Funding</a></li>
<li style="font-weight:400;">(00:29:26) <a href='https://www.theverge.com/openai/634802/openai-leadership-change'>OpenAI reshuffles leadership as Sam Altman pivots to technical focus</a></li>
<li style="font-weight:400;">(00:33:23) <a href='https://www.tomshardware.com/pc-components/gpus/nvidia-shows-off-rubin-ultra-with-600-000-watt-kyber-racks-and-infrastructure-coming-in-2027'>Nvidia shows off Rubin Ultra with 600,000-Watt Kyber racks and infrastructure, coming in 2027</a></li>
<li style="font-weight:400;">(00:35:23) <a href='https://asia.nikkei.com/Business/Tech/Semiconductors/China-s-SiCarrier-emerges-as-challenger-to-ASML-other-chip-tool-titans'>China's SiCarrier emerges as challenger to ASML, other chip tool titans</a></li>
<li style="font-weight:400;">(00:38:24) <a href='https://www.cnbc.com/2025/03/28/ponyai-wins-permit-for-fully-driverless-taxi-in-shenzhen-business-district.html'>Pony.ai wins first permit for fully driverless taxi operation in the center of China’s Silicon Valley</a></li>
</ul>
<ul>
<li style="font-weight:400;">Projects & Open Source</li>
</ul>
<ul>
<li style="font-weight:400;">(00:40:27) <a href='https://techcrunch.com/2025/03/24/a-new-challenging-agi-test-stumps-most-ai-models/'>A new, challenging AGI test stumps most AI models</a></li>
<li style="font-weight:400;">(00:45:16) <a href='https://arxiv.org/abs/2503.21380?utm_source=pocket_shared'>Challenging the Boundaries of Reasoning: An Olympiad-Level Math Benchmark for Large Language Models</a></li>
<li style="font-weight:400;">(00:48:13) <a href='https://arxiv.org/abs/2503.20314'>Wan: Open and Advanced Large-Scale Video Generative Models</a></li>
<li style="font-weight:400;">(00:50:38) <a href='https://www.artificialintelligence-news.com/news/deepseek-v3-0324-tops-non-reasoning-ai-models-open-source-first/'>DeepSeek V3-0324 tops non-reasoning AI models in open-source first</a></li>
<li style="font-weight:400;">(00:54:46) <a href='https://markdownlivepreview.com/nan'>OpenAI adopts rival Anthropic’s standard for connecting AI models to data</a></li>
</ul>
<ul>
<li style="font-weight:400;">Research & Advancements</li>
</ul>
<ul>
<li style="font-weight:400;">(00:55:56) <a href='https://www.technologyreview.com/2025/03/27/1113916/anthropic-can-now-track-the-bizarre-inner-workings-of-a-large-language-model/'>Anthropic can now track the bizarre inner workings of a large language model</a></li>
<li style="font-weight:400;">(01:06:00) <a href='https://arxiv.org/abs/2503.16779'>Chain-of-Tools: Utilizing Massive Unseen Tools in the CoT Reasoning of Frozen Language Models</a></li>
<li style="font-weight:400;">(01:11:50) <a href='https://arxiv.org/abs/2503.15299'>Inside-Out: Hidden Factual Knowledge in LLMs</a></li>
<li style="font-weight:400;">(01:15:14) <a href='https://sakana.ai/sudoku-bench/'>Sakana AI super-powers AI reasoning using Japan’s own Sudoku Puzzles</a></li>
</ul>
<ul>
<li style="font-weight:400;">Policy & Safety

</li>
<li style="font-weight:400;">(01:18:38) <a href='https://sd11.senate.ca.gov/news/senator-wiener-introduces-legislation-protect-ai-whistleblowers-boost-responsible-ai'>Senator Wiener Introduces Legislation to Protect AI Whistleblowers & Boost Responsible AI Development</a></li>
<li style="font-weight:400;">(01:21:50) <a href='https://wccftech.com/nvidia-other-tech-giants-demand-trump-administration-to-reconsider-ai-diffusion-policy/'>NVIDIA & Other Tech Giants Demand Trump Administration To Reconsider “AI Diffusion” Policy Which Is Set To Be Effective By May 15</a></li>
<li style="font-weight:400;">(01:23:17) <a href='https://www.cnbc.com/2025/03/26/us-blacklists-50-chinese-companies-in-bid-to-curb-beijings-ai-chip-capabilities.html'>U.S. blacklists over 50 Chinese companies in bid to curb Beijing's AI, chip capabilities</a></li>
<li style="font-weight:400;">(01:26:44) <a href='https://www.nytimes.com/2025/03/24/technology/reed-hastings-bowdoin-ai.html'>Netflix’s Reed Hastings Gives $50 Million to Bowdoin for A.I. Program</a></li>
</ul>
<ul>
<li style="font-weight:400;">(01:27:55) <a href='https://www.npr.org/2025/03/26/nx-s1-5288157/new-york-times-openai-copyright-case-goes-forward'>Judge allows 'New York Times' copyright case against OpenAI to go forward</a></li>
<li style="font-weight:400;">(01:29:48) <a href='https://www.avclub.com/umg-loses-injunction-anthropic'>Judge rules that AI can continue training on copyrighted lyrics, for now</a></li>
</ul>

Podurama

Home

Trending

Personal

Your Podcasts

New Episodes

My Files

Favourite Episodes

Tags (Folders)

Recently Played

Bookmarks

Followed Playlists

#205 - Gemini 2.5, ChatGPT Image Gen, Thoughts of LLMs

Last Week in AI

AI Chapters

AI Insights

Reviews