3 Minutes to reflect
If Dante relied on Virgil to navigate through hell, to understand something about the evolution of AI, my Virgil's name is Mario Alemi.
Thanks to him, I think I have understood several things. Some are absolutely counterintuitive.
Given the quality of ChatGPT's responses, we expect highly sophisticated algorithms, but in reality they are less complex than we think.
There have been no theoretical breakthroughs for years. The best performance is linked to the amount of available data (and partly to greater computational capacity).
A famous writing by Galileo Galilei describes mathematics as the language of the universe. The idea that with elegant formulas one can model the world is giving way to less elegant mathematics, but more accurate statistical correlations made possible by the gigantic amount of data available today.
In summary, the recipe for AI consists of 3 ingredients: (1) data + (2) statistical models + (3) computational capacity.
(2) and (3) are roughly commodities available to everyone. (1) is the differentiating element.
Once we said that data was the new oil: let's say that collectively we had seen the directional trend correctly, but as usual we didn't estimate the timing very well
2 Resources to pro level
Last week, Bloomberg launched BloombergGPT, which uses the same large language models as ChatGPT, but is trained on proprietary financial data.
I have a question for you: would you share your data in a world where data is the differentiating element or would you leverage your data as Bloomberg is doing?
Please meet my friend Mario, who is talking (in Italian unfortunately) about his latest project. He has built three AI-centric startups and has been researching the subject since 1990. He has also authored an amazing book on the subject, which is available on Amazon. Besides being an accomplished professional, he is also a wonderful human being and coffee connoisseur.
1 Reason to be happy
You can now ask generative AI to imagine scenes from your favorite books. This is Kafka on the Shore by Murakami. If you haven't read the book, this recommendation is a reason to be happy. And if you want one more reason, it turns out that what we call imagination is merely statistics. Well, maybe not a great reason to be happy after all.
If you liked this newsletter, share it via mail to a friend or on your socials. It would mean a lot to me.
Have a great weekend,
Simone
—
Want to read more: click here
Want to invest with me: click here
Want to share a deck: click here