DeepSeek
Unless you’ve been living under a rock for the past 24 hours, you’ll know by now about Deepseek, the new AI model released by new China AI startup. There are many winners and losers in the AI race, and many who have been placing their bets on specific companies may be surprised to see how recently their stocks have been plummeting.
Wisdom is learning the lessons we thought we already knew. DeepSeek reminds us of three important learnings from computing history:
— Pat Gelsinger (@PGelsinger) January 27, 2025
1) Computing obeys the gas law. Making it dramatically cheaper will expand the market for it. The markets are getting it wrong, this will make AI…
Julie Bort
https://finance.yahoo.com/news/deepseek-why-causing-investors-freak-145520325.html
I doubt it will be for too long, as this new open-source model will likely further advance the AI field for all competitors, not just Deepseek. Within months (or less), this code will already be part of, enhanced, and rebuilt into stronger models.
This is the beauty of open source—it’s an unstoppable force in tech development. Think of Unix descendants (semi closed source?), BSD, Linux, and projects like Tor, Android, and protocols like SMTP, and now Bitcoin, Nostr, etc. There are too many open-source projects to name them all here, but the point is that when a project releases its findings to the world, the world, in return, audits, forks, and builds on top of it.
OpenAI has been boasting about how much it costs to train AI models. Deepseek is a cold bucket of water for many right now (in terms of their bottom line and financial positioning), but in return, many will be advancing AI even further.
Regardless, lets keep the hype on computing power right? Again many players are taking a big hit on their bottom line.
but mostly we are excited to continue to execute on our research roadmap and believe more compute is more important now than ever before to succeed at our mission.
— Sam Altman (@sama) January 28, 2025
the world is going to want to use a LOT of ai, and really be quite amazed by the next gen models coming.
I’m not sure we are harnessing the power of AI in a wise way, and frankly, it’s frightening—on many different fronts.
Who is storing all this data? What are the costs of training? What is this data being used for?
On the other hand, we have this:
Feel free to change the flag of the CCP to the flag of the U.S. on the OpenAI Trojan horse. The amount of information we are giving to these companies and the terms we are accepting are absolutely out of control.
Before even testing it myself, I’d rather use something that isn’t coming from China in this case, after spending the last year studying the different cyberattacks that China has been behind.
I also don’t have the knowledge or capabilities to audit the code myself, so I’ll wait for the open community to figure out what’s really behind it. But I do look forward to the different versions and improvements this is going to bring.
Until the next one,
J