A Christmas AI Breakthrough: OpenAI’s o3 Passes the AGI Benchmark

In partnership with

Hi everyone,

This is Lore, the newsletter that helps you thrive in the age of AI.  

Thank you for being a part of 23k+ techno-optimists and growing. 🙏

I will keep this issue short and sweet because it’s Christmas Eve here in Japan on Tuesday when I’m writing this!

Next week, I’ll be taking off as the New Year is a big deal here. Also, I will be changing my schedule to release a newsletter issue every Thursday instead of Wednesday. So, I’ll see you again on January 8th!

P.S. Watch the latest episode of The Next Wave with Ammaar from Eleven Labs!

Your daily AI dose

Mindstream is your one-stop shop for all things AI.

How good are we? Well, we become only the second ever newsletter (after the Hustle) to be acquired by HubSpot. Our small team of writers works hard to put out the most enjoyable and informative newsletter on AI around.

It’s completely free, and you’ll get a bunch of free AI resources when you subscribe.

Click here to subscribe

1) Things Are Getting Better

2) OpenAI’s o3 Passes the AGI Benchmark

So, OpenAI saved the best for last. On the last day of their “12 Days of OpenAI,” they announced their upcoming o3 model (they skipped o2 for trademark reasons).

The announcement had me staring off into the distance for an entire day, thinking about what it means for humanity—what it means for you, me, and our families.

o3 is the first model to pass the ARC-AGI benchmark, an incredibly challenging test designed by the Alignment Research Center to detect artificial general intelligence–level reasoning, problem-solving, and self-improvement capabilities. (See the chart below.)

Of note: o1 Pro scores significantly higher than o1, which I’ve personally observed in my own testing of the model. It’s absolutely incredible how good it is at complicated problems like coding.

Just look at o3’s performance on competition-level math—phenomenal.

And here’s another chart showing how fast LLMs have improved at coding in just nine months: a stunning 71.7% improvement in less than a year.

o1-pro can already help you architect thousands of lines of code and often “one-shot” entire new feature additions—tasks that used to take teams weeks. What happens when AI can create entire products in minutes at our command?

And people have run IQ estimations, concluding o3 may clock in around 157 IQ (average is 100, 130+ is considered gifted, and Einstein was estimated at 160).

By late 2025, we’ll likely all have access to o3 or even o4—models at or beyond Einstein’s IQ, open for anyone to use.

To say this will change everything is an understatement. Those who believe it’s just hype aren’t thinking big enough about the second- and third-order effects of such powerful intelligence being widely accessible. Industries could change overnight, entire job categories might be disrupted, and innovation cycles could shrink from months to weeks—even days.

Meanwhile, o1-pro is already incredible. I’ve been using it to build a one-person game demo, which I showed to Matt on The Next Wave (the episode should drop in 1–2 weeks). He was blown away. Every aspect—art, video, code, story—was AI-generated, created part-time over two weeks while I did other things. This type of solo or two-person project is going to become normal. Entire companies, games, and even full-length movies will come from teams of just a few people by 2025.

Buckle up. This is going to redefine work, creativity, and, frankly, life as we know it. And all I can say is, here’s the best advice I’ve seen for how to handle the arrival of AGI:

3) AI Voice Technology Just Got INSANE

In this week’s episode of The Next Wave, Matt Wolfe and I had on our friend Ammaar from Eleven Labs. It’s amazing everything you can do with AI voice now. I highly suggest you check out this episode, it’s one of our best yet.

Also, please subscribe to The Next Wave on Apple, Spotify, and YouTube. 

4) Things I’m Learning From

Great short essay from Sam Altman about The Intelligence Age. When he wrote this in September, he would have already known about o1-pro and o3. Worth reading it again now that we know that we’ve basically achieved AGI and what we have now is likely the seed for ASI and will truly bring on The Age of Intelligence.

The latest interview with Sam Altman is worth watching. It sheds some more light on his complicated relationship with Elon Musk. And gives some hints for the future of OpenAI.

As always, Matt has the best weekly overview of all that has happened in AI. And wow, there was a lot this last week.

That’s all for today. Please consider sharing the newsletter with your friends if you think they’d enjoy it. 🙏

Please give me feedback on Twitter. I am still figuring out the best format for this newsletter. Which part did you like? Dislike? Do you have other suggestions? Please let me know—just tweet @NathanLands.

-Nathan