- Homescreen
- Posts
- 🏠 GPThink
🏠 GPThink
OpenAI is back
GM! It’s Brett again. Confession time: one of the worst things I did for my last company was getting caught up in the whole CEO performance act. Nearly tanked the whole operation.
Wrote down some signs to watch out for if you're heading down this slippery slope. Trust me, you'll want to avoid these startup killers.
Anyway, let's get to it.
BROUGHT TO YOU BY:

🏠 AI
o1

OpenAI dropped their latest model, o1, and it's got everyone buzzing. It is trained to chew on problems before spitting out answers, just like we do. It's learning to refine its thinking, try new angles, and even spot its own mistakes
here is o1, a series of our most capable and aligned models yet:
openai.com/index/learning…
o1 is still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it.
— Sam Altman (@sama)
5:29 PM • Sep 12, 2024
o1 basically thinks before it speaks, mimicking human problem-solving.
how it started how it's going 🍓
— edwin (@edwinarbus)
5:14 PM • Sep 12, 2024
It achieved 49th percentile / 213 points in this year’s competitive programming Olympiad (IOI) under human conditions of 50 submissions per problem. But with 10,000 submissions per problem, the model achieved a score of 362.14 — above the gold medal threshold.
OpenAI o1 — our first model trained with reinforcement learning to think hard about problems before answering. Extremely proud of the team!
This is a new paradigm with vast opportunity. This is evident quantitatively (eg reasoning metrics are already a step function improved)… x.com/i/web/status/1…
— Greg Brockman (@gdb)
6:19 PM • Sep 12, 2024
It scored 89% on Codeforces and even cracked the top 500 for AIME.
OpenAI’s o1 model takes 10-20s longer to reason before answering.
— Scores 89% on Codeforces, a popular competitive programming platform.
— Places top 500 for AIME, the qualifying math exam for the US Math Olympiad, the hardest national math exam for high schoolers!
— Deedy (@deedydas)
5:34 PM • Sep 12, 2024
Look at this o1's AIME accuracy. As test-time compute increases, o1's accuracy shoots. That’s wild.
Someone gave it the Norway Mensa IQ test and scored an IQ of around 120, potentially marking the first time an AI model has surpassed the average human IQ. That’s pretty WILD!
OpenAI's new o1 model is a BIG breakthrough in AI intelligence, if IQ tests say anything.
I gave it the Norway Mensa IQ test, and it blows other AIs out of the water.
I'm surprised!... Because there hadn't been public progress in the last 6mo.
Link to full analysis below:
— Maxim Lott (@maximlott)
7:50 PM • Sep 14, 2024
Right now, it's available for ChatGPT Plus users and top-tier API devs. They're planning to roll out a lite version to the free crowd soon.
The big picture: This is a significant leap in reasoning capabilities. We're talking about a model that's outperforming expert humans on PhD-level science questions. That's not just cool tech; it's a whole new ballgame for real-world applications. It's not AGI, but it's a step closer to AI that can truly think through complex problems.

Still struggling to achieve work-life balance and manage your time efficiently?
Join this 3 hour Intensive Workshop on AI tools & ChatGPT (usually $399) but FREE for first 100 readers.
🗓️ Tomorrow | ⏱️ 10 AM EST
An AI-powered professional will earn 10x more. 💰
An AI-powered founder will build & scale his company 10x faster 🚀
An AI-first company will grow 50x more! 📊
In this workshop, you will learn how to:
✅ Make smarter decisions based on data in seconds using AI
✅ Automate daily tasks and increase productivity & creativity
✅ Skyrocket your business growth by leveraging the power of AI
✅ Save 1000s of dollars by using ChatGPT to simplify complex problems

🤝 THE LATEST IN…
TECH
Japan is building the world's most powerful supercomputer.
OnlyFans creators made more money than all NBA players combined last year.
YC’s moving to 4 batches a year.
You can now post videos on Bluesky.
Bending Spoons plans to lay off 75% of WeTransfer staff after acquisition.
AI
Google rolling out Gemini Live to free users on Android.
White House launches an AI data center task force.
Apple’s planing to compete with Meta Ray-Ban.
DeepMind used Gemini 1.5 Pro to turn robots into helpful tour guides.
SCIENCE
NASA's Crew-9 mission gets a new launch date (Sept 25) after some shuffling – two astronauts instead of four, and a five-month ISS stay on the menu.
Greenland's mega-landslide made Earth vibrate for 9 days straight.
German startup Formo raised $61M to scale up its fungus-based "cheese".
California droughts might be fueling a nasty comeback for valley fever once the rain hits.
CRYPTO
DeFi yields set to pop as Fed eases up.
Bitcoin ETFs finally catch a break, raking in $263M in a single day.
Former FTX CEO Sam Bankman-Fried files appeal and requests for a new trial.
Bitcoin's playing under $60k while gold's hitting record highs.

🏃♀️ QUICKIES
Raise: Finally, an AI-powered bookkeeping and finance platform, raised $200M in a Series B funding round to expand its all-in-one financial services for SMBs.
Stat: $150B. That's the reported pre-money valuation OpenAI is seeking in talks to raise $6.5 billion
Rabbit Hole: The right kind of stubborn (Paul Graham)

🤩 MONDAY MOTIVATION

🛠️ FOUNDERS CORNER
The best resources we recently came across that will help you become a better founder, builder, or investor.
📈 Heap helps you with cross-platform product analytics.
📈 CapGo makes market research crazy fast.
🧠 AnimStats turn your boring stats into eye-grabbing gifs and videos.

❓️ AI GENERATED OR NOT

Is this AI generated? |

HOW WAS TODAY'S NEWSLETTER? |

REACH 40K+ FOUNDERS, INVESTORS & OPERATORS
If you’re interested in advertising with us, send an email over to [email protected] with the subject “Homescreen Ads”