Pinned post
Latest posts 5
Markets are perhaps the only viable test of superintelligence
They are easy to measure, hard to reward hack, and constantly evolving to evade automation
In the limit, understanding markets requires understanding the entire world
They are easy to measure, hard to reward hack, and constantly evolving to evade automation
In the limit, understanding markets requires understanding the entire world
The participation of AGI in capital markets is inevitable.View quoted post on X
Trading real, live financial markets is among the ultimate forms of truth seeking, and will likely become widely recognized as one of the most integral AI benchmarks.
There is perhaps no better venue for superintelligent price discovery than a mathematically-secured, global, transparent ledger of linear cash-settled perpetual derivative markets.
We would be happy to assist any efforts by @SpaceXSI (and others) toward this frontier
The biggest problems in financial markets x AI is the measurement problem
Markets are so complex that it's unclear what to measure and how to measure it accurately
A successful backtest could be a sign of a great strategy, but it probably isn't. More likely, its a sign of future leakage, overfitting, or luck
If you want to create a powerful learning flywheel, you need to solve this
It's part of why markets haven't "fallen" to RL yet, like code and math. In code and math, "good" is easier to disambiguate from luck or cheating
We've spent the majority of our time on this problem, and we finally made the necessary breakthroughs. You'll be able to feel the effects of this when we release our first family of models
Markets are so complex that it's unclear what to measure and how to measure it accurately
A successful backtest could be a sign of a great strategy, but it probably isn't. More likely, its a sign of future leakage, overfitting, or luck
If you want to create a powerful learning flywheel, you need to solve this
It's part of why markets haven't "fallen" to RL yet, like code and math. In code and math, "good" is easier to disambiguate from luck or cheating
We've spent the majority of our time on this problem, and we finally made the necessary breakthroughs. You'll be able to feel the effects of this when we release our first family of models
I just talked to the founder of a stealth neolab that closed a $1.2B seed round
Their initial plan is to build a time machine that allows them to travel multiple decades into the future and bring back ASI
Crazy times
Their initial plan is to build a time machine that allows them to travel multiple decades into the future and bring back ASI
Crazy times
AI labs and GPU providers are making lots of money selling tokens, but who's making money buying them?
In order to justify the next leg of token spend, companies have to reorganize around tokens as a driver of revenue vs. just increasing efficiency. This is especially true as models get more expensive & longer horizon.
There are some domains that are better positioned than others. Even in those domains, companies are just now starting to adapt. This post dives into the details
In order to justify the next leg of token spend, companies have to reorganize around tokens as a driver of revenue vs. just increasing efficiency. This is especially true as models get more expensive & longer horizon.
There are some domains that are better positioned than others. Even in those domains, companies are just now starting to adapt. This post dives into the details
View quoted post on X