Darren LI

2. If you are in the Pre-seed stage, the first thing is to go out to raise funds and prove the value of the group. In fact, some of the founders we work with have started lowering the valuations to find the funding they need over the next two to three years. In this case, you just need to raise money and go back to work.

4. This may be the best time to launch a product. There is too much noise in a bull market. As these startups face reality, they begin to understand that without the traction of the product, the token price will eventually drop and be blamed by the holders.

Particular product is bought because it was ranged and placed at eye level.

Many of these projects are saving time by training on small, highly curated datasets. This suggests there is some flexibility in data scaling laws. The existence of such datasets follows from the line of thinking in Data Doesn't Do What You Think, and they are rapidly becoming the standard way to do training outside Google

"It's fine to let the machine make all the decisions because they can do it at a much higher frequency and much more often to get the best possible results, but at the moment I don't have enough visibility into all the signals and where to spend my time improving which signals get the best-possible performance."

Our extensive test over 25 LLMs (including APIs and open-sourced models) shows that, while top commercial LLMs present a strong ability of acting as agents in complex environments, there is a significant disparity in performance between them and open-sourced competitors.