A Most Significant Week in AI

The AI Breakdown - Weekly Recap

Welcome back to The AI Breakdown, the most interesting & important news and conversations in AI. I noticed none of the AI newsletters seem to write over the weekend. Seems like someone should do that.

Product Launches & Updates:

• Midjourney v5.1 releases - a more “opinionated” version

• Stability AI’s DeepFloyd IF can actually put text in text-to-image creations

• Inflection releases more personal AI - Pi

• iBabyAGI puts AutoGPT in your pocket 

• Everyone is losing their mind over ChatGPT Code Interpreter 

• Ashton Kutcher spins up an oversubscribed $243m AI fund in 5 weeks

News from the Big Co’s:

• SlackGPT the latest tech co to go AI

• Bing Chat major upgrade including plugins + no more waitlist

• Microsoft-AMD AI chip rumors send AMD stock up

Industry-level Disruption:

• Samsung bans AI tools in workplace

• Meta warns of imposter ChatGPT malware

• ChatGPT cutting into subscriptions sends Chegg stock crashing 

• Disinformation much? Bloomberg finds 49 AI-created created content farms 

• The Writers Guild of America goes on strike, in part due to AI concerns

• IBM says 7800 planned jobs won’t be hired due to AI

  • (which kinda makes this 👇 argument a little…less convincing?)

AI Regulation & Geopolitics:

• White House announces slate of AI initiatives, including $140m in funding and new public review for models

• New ‘AI Czar’ Kamala Harris meets with AI CEOs

• ChatGPT returns to Italy

• G7 meeting focuses on AI

• Just a week after the Chinese politburo encouraged the development of AGI, Chinese president Xi Jinping calls for more focus on AI

AI Safety Goes Mainstream:

• Warren Buffett compares AI to the atomic bomb: “We won’t be able to un-invent it”

• AI pioneer Geoffrey Hinton left Google and has been on a media tour warning about the dangers of AI - interestingly, it’s moved that conversation more to the mainstream than we’ve seen with previous warnings

Open Source: Outcompeting Google?

• HuggingFace releases an open source alternative to Github’s Copilot

• Replit announces generative coding language model ReplitLM

• This leaked memo from a Google research on why they (and OpenAI, and any other centralized companies) are going to be out-competed by open source. Summed up below

The Latest Research

• A GPT-style model trained directly on individuals’ brains using fMRI was able to accurately produce text from thoughts — IRL mind reading!

• An AI that converts brain signals to video? CEBRA.ai publishes in Nature

This Week in the AI Breakdown

• Mon May 1 | Text-to-Image AI That Can Actually Spell!? Meet DeepFloyd IF

• Tue May 2 | Mind-Reading AI? Scientists Use GPT-Like AI to Accurately Translate Thoughts into Text

• Wed May 3 | The First Anti-AI Strike: The Writers Guild of America Protests AI Writing

• Thu May 4 | 6 Ways ChatGPT Code Interpreter Is Already Being Used

• Fri May 5 | Google Losing to Open Source AI? "We Have No Moat, And Neither Does OpenAI"

• Sat May 6 | AI Just Quietly Had One of It's Most Important Weeks in Recent History

Share The AI Breakdown, win free Midjourney or ChatGPT accounts!

We so appreciate all you taking the time to read The AI Breakdown. We’ve just added new referral rewards.

• If you refer just 1 other subscriber you’ll be entered to win a weekly drawing for a free year of Midjourney’s base plan

• If you refer 2 other subscribers you’ll also be entered to win a weekly drawing for a free 6-months of ChatGPT Plus

Thanks for reading!