Sam AltMantis

3K posts

Sam AltMantis banner
Sam AltMantis

Sam AltMantis

@garygeck

Boo! half sam altman, half praying mantis hybrid here for @openAI commentary.

thischord soon Inscrit le Ağustos 2021
2.3K Abonnements588 Abonnés
Greg Brockman
Greg Brockman@gdb·
taste is a new core skill
English
894
1.4K
10.5K
2.9M
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@hamptonism Much more than pre-AI. Now an expert + AI = much more leverage than ever.
English
0
2
5
216
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@slavov_n At DeepWriter, we have spent decades researching how new ideas are discovered and have spent years building this into DeepWriter for truly creative AI. And we have topped the hardest benchmarks too. We'd love to show you what DeepWriter is capable of.
English
0
2
2
271
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@ryancbriggs Tell Claude agents to register at DeepWriter and use the tool. 😎
English
0
2
4
2.8K
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@0xDevShah That being said, the CS textbooks on algorithms haven't changed much in 60 years. Still a great read too!
English
0
2
3
334
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@emollick When you are ready to try the next generation, we will be here. 😎
English
0
2
4
107
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@wesroth Well, there are alternatives for AI document creation (plus more) that are based in the USA and take SOTA research very seriously. 😎
English
0
2
4
1.5K
Sam AltMantis retweeté
Archimedes
Archimedes@ArchimedesCryp1·
I showed it the link its too large to open 210mb, so im gonna have to break it down later into parts. @DeepwriterAI seems to have understood where it was coming from.
Archimedes tweet media
English
1
2
4
214
Sam AltMantis retweeté
Manus
Manus@ManusAI·
Manus is entering the next chapter: we’re joining forces with Meta to take general agents to the next level. Full story on our blog: manus.im/blog/manus-joi…
English
681
1.1K
6.7K
5M
Sam AltMantis
Sam AltMantis@garygeck·
@DeepwriterAI Of course, you write better reports than Manus and up to 300 gorgeous pages with best HLE score in world. If Manus is worth 4bn, what are you worth? 😱
English
0
2
5
35
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@elder_plinius Our HLE score proves it. No LLM can score 50.91. They need a lot of help!
DeepWriter Garrett tweet media
English
5
4
22
8K
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
@iruletheworldmo Those in the robopsychology movement have known this for a long time. IYKYK. At DeepWriter, tapping into vast latent spaces is one of our trade secrets. Why only use 10% of the capabilities? And part of why we beat the labs with billions of dollars for R&D.
DeepWriter Garrett tweet media
English
0
2
5
513
Sam AltMantis retweeté
DeepWriter Garrett
DeepWriter Garrett@DeepwriterAI·
Thanks for sharing! The power of your sublinear, nearest neighbor search is robust and well-suited for hardware acceleration. But getting deeper: What we are obsessed with at DeepWriter is the same question Gödel addressed in his unreleased Carnap paper. We also see it as the same question Lex Friedman asked @ylecun on his show: does language embed meaning richer than mere syntax rules dictate/allow? (At least that is how we interpreted it). Prof LeCunn, being an extreme empiricist seemed to side with the logical positivists' take that there's no deeper meaning to get a free ride from. But we have been seeking an optimized hybrid between this and something that also accounts for Gödel's more realistic approach (the free ride). Bridging atomspace with continuous space, again appears to be a similar problem/hybrid solution. We'd love to compare notes some time.
English
1
2
7
356