Last Week in AI· Folge 202

#163 - Llama 3, Grok-1.5 Vision, new Atlas robot, RHO-1, Medium ban

Donnerstag, 25. April 2024 · 1 h 34

MP3

Shownotes

Our 163rd episode with a summary and discussion of last week's big AI news!

Note: apology for this one coming out a few days late, got delayed in editing it -Andrey

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

Email us your questions and feedback at contact@lastweekin.aiand/or hello@gladstone.ai

Timestamps + links:

• Intro / Banter

• Tools & Apps

• () Meta releases Llama 3, claims it’s among the best open models available (techcrunch.com/2024/04/18/meta-releases-llama-…)

• () Elon Musk’s xAI Unveils Grok-1.5 Vision, Beats OpenAI’s GPT-4V (analyticsindiamag.com/elon-musks-xai-unveils-g…)

• () Reka releases Reka Core, its multimodal language model to rival GPT-4 and Claude 3 Opus (venturebeat.com/ai/reka-releases-reka-core-its…)

• () Cohere Compass Private Beta: A New Multi-Aspect Embedding Model (txt.cohere.com/compass-beta)

• () Amazon Music’s Maestro lets listeners make AI playlists (theverge.com/2024/4/16/24132129/amazon-music-m…)

• () Snap plans to add watermarks to images created with its AI-powered tools (techcrunch.com/2024/04/17/snap-plans-to-add-wa…)

• Applications & Business

• () Boston Dynamics unveils new Atlas robot for commercial use (nbcnews.com/news/weird-news/boston-dynamics-un…)

• () TSMC’s $65 billion bet still leaves US missing piece of chip puzzle (arstechnica.com/gadgets/2024/04/tsmcs-65-billi…)

• () U.S. blacklists Intel's and Nvidia's key partner in China — three other Chinese firms also included in the blacklist for helping the military (tomshardware.com/tech-industry/semiconductors/…)

• () Elon Musk says the next-generation Grok 3 model will require 100,000 Nvidia H100 GPUs to train (tomshardware.com/tech-industry/artificial-inte…)

• () Dr. Andrew Ng appointed to Amazon’s Board of Directors (aboutamazon.com/news/company-news/dr-andrew-ng…)

• () Collaborative Robotics Locks Up $100M, Latest Robot Startup To Raise Big (news.crunchbase.com/robotics/startup-collabora…)

• Projects & Open Source

• () OpenEQA: Embodied Question Answering in the Era of Foundation Models (docs.google.com/document/d/1_x7RzO0YTpLPaigeix…)

• () Introducing Idefics2: A Powerful 8B Vision-Language Model for the community (huggingface.co/blog/idefics2)

• Research & Advancements

• () RHO-1: Not All Tokens Are What You Need (arxiv.org/abs/2404.07965)

• () Scaling Laws for Fine-Grained Mixture of Experts (arxiv.org/abs/2402.07871v1)

• () Chinchilla Scaling: A replication attempt (arxiv.org/abs/2404.10102)

• () China develops new light-based chiplet that could power artificial general intelligence — where AI is smarter than humans (livescience.com/technology/electronics/china-d…)

• () OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments (arxiv.org/abs/2404.07972v1)

• Policy & Safety

• () U.S. Commerce Secretary Gina Raimondo Announces Expansion of U.S. AI Safety Institute Leadership Team (commerce.gov/news/press-releases/2024/04/us-co…)

• () NSA Publishes Guidance for Strengthening AI System Security (nsa.gov/Press-Room/Press-Releases-Statements/P…)

• () Foundational Challenges in Assuring Alignment and Safety of Large Language Models (llm-safety-challenges.github.io/challenges_llm…)

• () Former OpenAI Board Member Calls for Audits of Top AI Companies (bloomberg.com/news/articles/2024-04-16/former-…)

• () SoA survey reveals a third of translators and quarter of illustrators losing work to AI (www2.societyofauthors.org/2024/04/11/soa-surve…)

• Synthetic Media & Art

• () Medium bans AI-generated content from its paid Partner Program (bleepingcomputer.com/news/technology/medium-ba…)

See Privacy Policy at art19.com/privacy and California Privacy Notice at art19.com/privacy#do-not-sell-my-info.

0:00 / 0:00