AI in a Nutshell - Week 43 - Atlas Browsers, Banned Superintelligence, and AI Doing Homework
OpenAI drops the ChatGPT Atlas browser, Anthropic gets a million TPUs from Google, Elon Musk predicts jobs will become optional, and 850 tech leaders call for a pause on superintelligence.
Fellow human, it’s week 43. AI hasn’t taken over the world yet. And here’s your lazy-man version of what went down in the AI world in the past week.
Speed-read: OpenAI drops the ChatGPT Atlas browser, Anthropic gets a million TPUs from Google, Elon Musk predicts jobs will become optional, 850 tech leaders call for a pause on superintelligence, Apple launches the M5 chip, and Karpathy still swears AGI is a decade away.
What You Must Know
Note: You can click the title to read more on the subject.
OpenAI Drops ChatGPT Atlas Browser → OpenAI launched Atlas, an AI-powered browser. It’s like Chrome with a built-in AI assistant. I used it briefly and wrote about my first impressions here, you should check it out.
Elon Says AI Will Make Jobs Optional → Musk claims AI and robots will make all jobs optional. What came to my mind was “Will bills become optional as well? 😒”.
850 Tech Leaders Demand a Ban on Superintelligent AI → Over 850 global tech figures signed an open letter calling for a pause on developing “superintelligent AI” until global safety standards exist. Isn’t it quite late to stop now? After OpenAI and co. already invested billions? 🤔 It’s like yelling “Stop the car!” after driving off the cliff.
China Unveils Its National AI Strategy → China released a new AI plan linking model scaling to renewable energy capacity. It’s a smarter shift from the Western “bigger model = better AI” mindset. It’s part tech policy, part energy plan, and might become a template for other countries aiming to balance power and progress.
AGI in 10 Years? Karpathy Still Thinks So → Andrej Karpathy, former Sr. Director of AI at Tesla, said hireable digital workers (not just robots) will be with us within a decade. Now, we wait.
What’s Good to Know
Note: You can click the title to read more on the subject.
ChatGPT Projects Now Shareable for Free Users → Shared Projects are now open to everyone, not just paid users. I always thought this was one of its most underrated ChatGPT features, glad it’s finally mainstream.
NanoChat Gets a Personality Update → Andrej Karpathy’s $800 homegrown AI, nanochat, now has a self-aware “personality”. It talks about its creator, knows its limits, and jokingly calls him “King Andrej”. It’s a fun example of how smaller models can still feel surprisingly personal.
Yann LeCun Says AI Doom Talk Is Overblown → Meta’s AI chief, Yann LeCun, says everyone should calm the fan down. AI is not a bubble. His take: “risks are real but speculative. We’ve handled bigger technologies before.”
Google Veo 3.1 Challenges Sora → Google’s new Veo 3.1 upgrade lets creators extend clips, refine scenes, and generate smooth transitions; it’s now arguably the best choice for AI video creation.
Debate: Do AI Browsers Actually Matter? → TechCrunch debated whether AI browsers like Atlas or Comet are actually useful. The consensus: small productivity boost, big implications for the open web. If they succeed, the way we “browse” might change completely.
AI Tools Worth Knowing
BrowsOS → An open-source agentic browser.
CalPulse → AI menu scanner that helps you decide what to eat.
Kodus → AI-powered code reviewer.
HiFriday → An AI that calls you for self-reflection.
Amigo AI → Your personal finance buddy powered by AI.
Kaizen Corner → Learning AI, the DeepMind Way
DeepMind teamed up with University College London to launch a free learning hub for anyone who wants to actually understand AI, not just use it.
I checked it out. It’s a bit technical, but the lessons are bite-sized and surprisingly chill. If you’ve ever said, “I want to learn AI but don’t know where to start,” this might be a moderately painful option.
Meme of the Week
That’s your week in AI.
If you learned something, tell a friend. And if you didn’t, blame yourself.
Until next Sunday,
Kay - your fellow human
P.S. If this email lands in spam, that’s your inbox trying to stop you from staying plugged in. Fix it.