- AI Fire
- Posts
- 🌀 'ChatGPT Psychosis' is So Real
Read time: 5 minutes
AGI? That was last year’s buzzword. Now the real race is to build superintelligence by 2027. Meta’s rebranded, Altman’s hinting at it, and the U.S. might drop $244B.
What are on FIRE 🔥
LEARNING PARTNER AIRCAMPUS
You don’t need another productivity hack. You need a system that works without you.
While most people juggle tabs and chase tasks, a few are handing it all off - to AI agents that don’t assist, they act.
Imagine waking up with your inbox cleared, content posted, and calendar perfectly managed - all handled quietly, without you lifting a finger.
On 8th July at 10 AM EST, we're hosting a Masterclass showing you how to architect AI agents that actually run your work - across 40+ tools - so you can focus on the parts only you can do.
If you're ready to step into the next version of how you work, now’s the moment.
🎟️ Save Your Seat — and let the systems take it from here.
AI INSIGHTS
AGI? As in, artificial “general” intelligence? That's so 2024. Now, everyone's racing to build artificial “superintelligence” by 2027. Just look at the marketing language:
Meta renaming its entire lab to Meta Superintelligence Labs
Sam Altman calling our current moment a “gentle singularity”
Anthropic predicting a “country of geniuses in a data center” by 2026–2027
There’s the theory AGI = “A Genius Indian”, ASI = “A Soham Indian” lol. In a new interview, the OpenAI investor and tech oracle dropped this timeline:
2025-2030: The AI Intern Era: Every pro gets an AI co-worker. AI systems smarter than Stanford grads handle 80% of your job
2030s: Corporate Extinction Event: Fortune 500 collapse. Tiny teams build billion-dollar companies. AI “interns” surpass human bosses
2040+: Work Is Optional: You work for passion, not rent. Robots handle all physical labor. Education, healthcare, legal advice = free
He’s 80% confident this is where it’s going. So are we already there? Sort of. We’re in what some call the Jagged AGI era:
OpenAI o3: Can generate entire startup plans and crush AGI benchmarks… but fails at riddles.
Google AlphaFold: Predicts protein structure with atomic precision… but struggles with broader chemistry tasks.
So yeah, we might already have AGI-level performance in certain domains. Which means… superintelligence is now the only milestone that matters.
Why It Matters: Superintelligence is now a company strategy, a hiring pitch, a research North Star, and a fundraising buzzword. We’ve officially entered the “build or die” phase of AI. And no one’s pretending otherwise.
TODAY IN AI
AI HIGHLIGHTS
🤐 Musk’s own AI just blames him (& Trump) for recent deadly Texas floods, too woke. It doesn’t care if it’s left or right. If the data links you to disaster, it’ll call it out.
🧩 If you're still waiting on DeepSeek R2, you’ll love this report: DeepSeek vs. the World. It compares all about this model with others. Things are more complicated.
😵 Managers are letting ChatGPT decide who gets fired… and yeah, it’s as bad as it sounds. You might lose your jobs because a chatbot said so! So 'ChatGPT psychosis'.
🕵️♂️ Turns out, researchers are sneaking white-text prompts into preprint papers to whisper sweet nothings to AI reviewers using tiny fonts. This turns peer review into SEO warfare.
🧬 Wow, Google DeepMind truly wants to cure all diseases… and they're (finally) testing drugs on humans with real pharma experts. See its progress here.
💰 AI Daily Fundraising: Genesis AI has secured $105 million in seed funding from Eclipse Ventures and Eric Schmidt, to develop AI for automating the $30-40 trillion global physical labor market.
AI SOURCES FROM AI FIRE
NEW EMPOWERED AI TOOLS
🧠 Teammates.ai is an autonomous AI Interviewer that screens thousands of candidates/hour, for any role in 50+ languages. You’ll get a full insight-rich report in minutes. Just upload your JD.
🗺️ AI Mindmap transforms pages, PDFs, YT videos into mind maps instantly.
💸 Cloudflare Pay Per Crawl charges AI crawlers for accessing your page.
✍️ SEO Content Lead is a SEO-optimized content for ChatGPT in seconds.
AI QUICK HITS
🔥 Here's one of the best coding interviews with AI in 500–1000 interviews.
🛰️ Astronomers saw the weirdest signals ever recorded. It’s extraterrestrial…
😰 Improved Grok criticizes Democrats and Hollywood’s ‘Jewish executives’.
🧠 A research reveals that in 2024, AI helped write 13.5% of scientific papers.
🔴 Google faces EU antitrust complaint over AI Overviews from publishers.
AI CHART
The buzz around a US-led “AI Manhattan Project” is starting to look logistically feasible. And if it happens, it dwarfs anything we’ve seen so far, 10,000x GPT-4 by 2027 kind of big.
What is an “AI Manhattan Project”? People throw around the term a lot, but here’s how they define it:
Government-led: Initiated and backed by the US government.
Private compute consolidation: National-scale coordination with companies like NVIDIA, OpenAI, etc.
Massive GDP investment: Spending reaches Apollo/Manhattan levels, up to 0.8% of US GDP (~$244B/year).
That’s not theoretical. The US-China Economic and Security Review Commission recommended something like this to Congress. And even the Dept. of Energy is tweeting about it.
If all those are used together for 100 days, you get ~3e29 FLOP, double the previous estimate, and still plausible with current US energy and hardware trends.
Just by consolidating existing US compute resources, you get a 1-year head start on scaling curves. They project this setup would deliver compute on par with 2028-level systems, by the end of 2027.
So it’s a moonshot, but the infrastructure and money already exist. It just needs political will and coordination.
We read your emails, comments, and poll replies daily
How would you rate today’s newsletter?Your feedback helps us create the best newsletter possible |
Hit reply and say Hello – we'd love to hear from you!
Like what you're reading? Forward it to friends, and they can sign up here.
Cheers,
The AI Fire Team
Reply