Episode Details
| Title | #50 – Alexander Long on Pluralis Research and protocol learning for frontier models |
|---|---|
| Date | 2/25/2025 |
| Podcast | Into the Bytecode ↗ |
Key Insights
Bullish:
- 🟢 $DECENTRALIZEDAI - Truly open-source, decentralized AI development is expected to outpace centralized efforts in the long run, drawing parallels from historical trends in open-source software development.
- “I think there is no actual open source, no true open source at the model layer today. So we talked about it at the start. I think these open weight models, it's not open source. And my view is historically, if you ever have genuine community engagement and like you have this, this, almost this, the, the base trunk of the tree, that's just pushing everything forward. It doesn't matter. When a when a centralized company branches off, they can they can get ahead for a little bit, but that main trunk of progress that has all the research retention, all the academic attention, all the effort that's going there because people know it's truly open, it always, ends up moving faster in the end.” - Alexander (51:19)
- 🟢 $AITOOLS - 🔒
- 🟢 $SMALLAICOMPANIES - 🔒
- 🟢 $SOFTWAREDEV - 🔒
Bearish:
- 🔻 $KNOWLEDGEWORK - 🔒