Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
gamerpuppy
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
gamerpuppy
2y ago
Location: Seattle Remote: Yes. Hybrid 1-2 days possible. Willing to relocate: No. Technologies: C++, Rust, Java, system design / optimization, organizational improvement Resume: https://docs.google.com/document/d&#
2.
▲
by
gamerpuppy
2y ago
I really dislike holding up scientists as a type of person to aspire to. Science is just a tool, and is frequently used to damage the interest of the public good. I would rather venerate those who are veritably kind to their friends and com
3.
▲
by
gamerpuppy
3y ago
LLm's are agi, you can use them for anything so posts like "LLm's for x" are really so boring.
4.
▲
by
gamerpuppy
3y ago
I could if I had strongly supportive cofounder… basically romantic partner level of commitment and enough time.
5.
▲
LLMs End Nuclear Peace
(twitter.com)
1 points
by
gamerpuppy
4y ago
|
0 comments
6.
▲
The current and future state of AI/ML is shockingly demoralizing (2022)
(old.reddit.com)
5 points
by
gamerpuppy
4y ago
|
0 comments
7.
▲
by
gamerpuppy
4y ago
Yes it is obviously not an AGI in the sense of an intelligent, persistent agent, but it is also obviously a huge step towards one. It's like a single pass of thought on a topic, combined with self iteration and recursion in answer gene
8.
▲
by
gamerpuppy
4y ago
ChatGPT is general AI. It perform actions in a class requiring abstract thought which previously only humans were capable of. Sure the applications we see it is capable of are limited now, but that’s a consequence only of it’s operating en
9.
▲
by
gamerpuppy
4y ago
We need to start talking about AI safety and nationalizing AI research like yesterday. ChatGPT is already having destabilizing effects and it is frightening how little we are discussing (this hugely powerful tool)'s potential as a weap