respectattentio

4 ms·

31.
by respectattentio 2mo ago
is it a benchmarkmaxxing model?!
32.
by respectattentio 2mo ago
As simple as a prompting it with structural output or restrictions for your criteria. With agents, the prompting could be dynamic for maximum accuracy for every retrieval. This absolutely would beat the best of the best embedding-based RAG
33.
by respectattentio 2mo ago
You have a point! It appears bigger. That's due to the compound effect I guess (more internet users -> more data -> better models -> fast repeat). Even though 01 -> asm was revolutionary, but its loop was not as big and as
34.
by respectattentio 2mo ago
I believe embedding-based RAG, everybody is using, will end. As chips advance, you would use a big llm instead of word embedding for retrieval. It's much more accurate and extensive covering every topic. Still need ~2 years to be repla
35.
One MCP directory for all docs.domain.tld (No auth needed)
3 points by respectattentio 2mo ago|0 comments
36.
by respectattentio 2mo ago
It's the one who understands the universe and how it works wins. Not who got the first idea Not who built it first Not even in the middle Any doubt? check Nikola Tesla's life. I believe this aligns with what Peter Thiel said: &quo
37.
by respectattentio 2mo ago
Telegram CEO predicted this the moment countries started to announce that. His idea, that I completely agree with, is: Awareness > Control
38.
by respectattentio 2mo ago
programming has changed and leveled up due to ai it's as big as going from zeros and ones to assembly language
39.
by respectattentio 2mo ago
it's for sure better than deepseek flash 07/31
40.
by respectattentio 2mo ago
they are playing a completely different game than the US
41.
by respectattentio 2mo ago
I came up with this after building a few systems. For me, if I put costs in my architecture, I'm limited heavily and that can easily change how memory is shaped dramatically. The opposite, putting memory in architecture, is not true. M
42.
by respectattentio 2mo ago
I like to start with memory engineering then reach full system then reducing costs. This allows unlocking full potential of agents.
43.
by respectattentio 2mo ago
science?!! but I'm working to fix that...
44.
by respectattentio 2mo ago
Speed, security, or scaling — choose 3 with our hosted version: https://outbidlaunch.com
45.
Show HN: OutbidLaunch – Open-Source Pay-to-Rank Leaderboard with 3D Live Globe(github.com)
1 points by respectattentio 2mo ago|1 comments
46.
AI isn't the end goal. (my weird view)
2 points by respectattentio 2mo ago|0 comments