@wedtm It seems they are the ones providing the liquidity needed to move the market. Without them, even a token with a strong project may struggle to gain traction because most buyers want to see momentum and trading activity before getting involved
One thing I took away from this debate is that AI adoption is still in its relatively early stages globally.
This is why projects like @SentientAGI may be ahead of the curve, especially with their focus on open source and decentralized AI.
As AI becomes more widely adopted, transparency and user control could become increasingly important, which may lead many people to prefer open source AI over closed source alternatives.
A lot of complex AI agents use RAG for tool filtering or prompt selection, yet there is very little public information about how it works in practice. @SentientAGI decided to address that by sharing insights from their experiments.
Sentient findings show that when agents have large toolsets, RAG-based filtering can improve tool selection, reduce failures and lower latency. Combined with parallel execution, it helps agents solve queries more efficiently without sacrificing answer quality.
Lots of complex production agents actually use RAG for tool filtering or prompt selection, and yet there is barely any information about it. We decided to fix that.
When your agent has 40+ tools, brute-force selection breaks down. We measured how embedding-based filtering and parallel execution cut failures by 29% and latency by 40% — and why 95% of queries now complete in ≤2 iterations.
Read more below ↓
Lots of complex production agents actually use RAG for tool filtering or prompt selection, and yet there is barely any information about it. We decided to fix that.
When your agent has 40+ tools, brute-force selection breaks down. We measured how embedding-based filtering and parallel execution cut failures by 29% and latency by 40% — and why 95% of queries now complete in ≤2 iterations.
Read more below ↓