- The Rundown AI
- Posts
- OpenAI co-founder quits (again)
OpenAI co-founder quits (again)
PLUS: Apple's mysterious AI domain purchase
Rowan Cheung
February 15, 2024
Welcome, AI enthusiasts.
OpenAI’s research team just took a major blow, with co-founder and renowned AI researcher Andrej Karpathy announcing his departure from the company.
With this exit and the lingering mystery surrounding Ilya Sutskever’s exile — is something bigger brewing behind the scenes at OpenAI? Let’s investigate…
In today’s AI rundown:
Andrej Karpathy departs OpenAI
Apple makes mysterious AI domain purchase
How to save time & money when shopping using Vetted
LLMs improve human forecasting accuracy
7 new AI tools & 4 new AI jobs
More AI & tech news
Read time: 4 minutes
LATEST DEVELOPMENTS
OPENAI
Image source: Getty Images
The Rundown: Renowned AI researcher and founding OpenAI member Andrej Karpathy just announced he is departing the company for a second time, posting on X that he plans to pursue personal projects.
The details:
Karpathy helped found OpenAI in 2016 before serving as Tesla's Senior Director of AI for five years, rejoining Sam Altman and co. in 2023.
The departure follows OpenAI’s drama with Altman and the board, which has since relegated fellow co-founder Ilya Sutskever to an unclear role in the company.
Karpathy emphasized that no drama or event led to the departure and that he was simply pursuing personal projects (including his famous AI lectures on YouTube).
Why it matters: Between the mystery surrounding Ilya Sutskever and now Karpathy’s departure, something seems fishy between the science and business arms of OpenAI. All eyes now turn to where Karpathy heads next — with likely no shortage of suitors for one of the world’s top AI minds.
TOGETHER WITH IMAGINE AI LIVE
The Rundown: Join us at the Imagine AI Live conference, March 27-28, at the new Fontainebleau Las Vegas, and discover how AI is transforming the business landscape.
Learn from industry experts, and network with like-minded professionals.
Discover how AI can speed up your enterprise, expand your business model, and improve your value proposition
Spots are filling up fast! Don't miss out on this chance to future-proof your business. Sign up before February 15 and save 15%.
APPLE
Image source: Apple Insider
The Rundown: Apple just acquired a new domain shortly after open-sourcing a new image editing model and releasing several AI-focused research papers, sparking speculation of major AI plans for its productivity apps.
The details:
Apple recently purchased the rights to iWork.ai at an expired domain auction, coinciding with expected major AI moves for Apple’s suite of products.
iWork has been used in Apple branding for nearly 20 years, referring to the suite of OS apps, including Keynote, Pages, and Numbers
Apple recently released MLLM-Guided Image Editing, and reports have claimed iOS 18 will highlight Apple's biggest AI leap yet.
Why it matters: Between MGIE's release, new AI research, and the iOS 18 rumors, Apple appears to finally be building momentum in the sector after a slow start. But this domain purchase making headlines could be the rumor mill trying to will new AI moves into existence. Patience, Apple lovers!
AI TRAINING
The Rundown: Vetted built an AI shopping agent using their own fine-tuned LLMs & product data. It does extensive, real-time research to give users trustworthy shopping advice.
Here is a quick step-by-step:
Download the Vetted app for free & just ask
Get recommendations based on consensus from Reddit, user reviews, YouTube & blogs
Spot key differences, know what to consider, and get well-researched answers to follow-up questions
See prices across stores at a glance and never overpay again
PRESENTED BY SCRIBE
The Rundown: Scribe just bagged $25M in Series B funding to give you a well-deserved break from answering people’s questions all day. Instead, just create step-by-step guides (automatically, thanks to AI) to share with your team.
Capture any process using the Chrome extension
Easily customize steps and redact sensitive info
Share with colleagues – and get back to your work
AI RESEARCH
Image source: Midjourney
The Rundown: New research just found that access to a GPT-4T LLM assistant boosted forecasting accuracy by 23% over basic AI access, regardless of the assistant's bias — indicating benefit beyond prediction accuracy.
The details:
The study tested both biased (instructed to give extreme forecasts) and unbiased (instructed to be an objective ‘superforecaster’) assistants.
Both biased and unbiased LLM assistants helped humans make much better forecasts.
Excluding one confusing question, the unbiased assistant improved human accuracy by 43%, while the biased one improved it by 28%.
Why it matters: This study focused on predictions, but the ability of even biased AI to enhance human reasoning likely applies to many judgment-heavy areas like business, policymaking, and medicine. One day soon, we may be bouncing ideas off a panel of super-intelligent assistants to analyze a decision.
NEW TOOLS & JOBS
📅 Broadcast- Capture meeting notes, automate action items, and boost productivity (link)
🔑 Fixkey- A Grammarly alternative for macOS power users (link)
🎨 Visme- Create stunning visual designs effortlessly with AI assistance (link)
🌍 Panoramai- Transform Google Street View into your personalized world (link)
📝 NotesGPT- Transform voice notes into clear and concise action items effortlessly (link)
🛍️ Shop That Look- Upload a photo of an outfit you like, and find stores that offer similar items (link)
🤖 OS-Copilot- Towards Generalist Computer Agents with Self-Improvement (link)
QUICK HITS
Google is establishing a new AI hub in France, with plans to host around 300 researchers and engineers in Paris following its inauguration this week.
Stability AI introduced a research preview of Stable Cascade, a new image generation model with a new architecture for more efficient and detailed image creation.
A California court partially dismissed the copyright case against OpenAI, brought by authors including comedian Sarah Silverman, keeping the direct infringement claim while throwing out others.
Replit introduced new collaborative features in AI Chat, allowing for joint coding sessions, as well as the ability to manage multiple chat sessions and seamlessly switch between tasks without losing progress.
Google has been bolstering its fight against fake online reviews using an AI algorithm, blocking over 170M fraudulent reviews in 2023 to protect businesses on Maps and Search.
Artist Alicia Framis is set to become the first woman to marry an AI-generated hologram, as part of her art project 'Hybrid Couple’.
OpenAI and Microsoft have disrupted operations of five state-affiliated cyber threats, including malicious actors out of China, Iran, North Korea, and Russia.
THAT’S A WRAP
SPONSOR US
Get your product in front of over 450k+ AI enthusiasts
Our newsletter is read by thousands of tech professionals, investors, engineers, managers, and business owners around the world. Get in touch today.
FEEDBACK
How would you rate today's newsletter?Vote below to help us improve the newsletter for you. |
If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.
Join the conversation