Davidzheng

5 ms·

91.
by Davidzheng 4mo ago
Maybe what can happen is that the people who benefit from it donate back into the organization of this room when they have the means decades later?
92.
by Davidzheng 4mo ago
ok I guess they had other clues then if you do any sort of comparison vs Nex & Qwen probably a lot of weird coincidences will show up if somehow the three weights are not linearly independent lol
93.
by Davidzheng 4mo ago
it's interesting that this was even guessed at
94.
by Davidzheng 4mo ago
If they only opposed it they wouldn't have had these billions of dollars? Also I think they genuinely believe they cannot stop it because the Chinese companies are close behind (I also believe it's impossible to stop b/c of s
95.
by Davidzheng 4mo ago
Are you sure most of frontier cost isn't inference in RL environments?
96.
by Davidzheng 4mo ago
How is it ironic to believe in a risk that actualizes
97.
by Davidzheng 4mo ago
Theres quite a large number of people who believe it's basically impossible when the intelligence gap is too big
98.
by Davidzheng 4mo ago
I agree with this. But i think Ilya and Dario hold these beliefs sincerely. Probably a sizable portion of Anthropic employees too
99.
by Davidzheng 4mo ago
We're on track to get there globally and economic pressures will ensure it happens. It's not too early to worry about it
100.
by Davidzheng 4mo ago
Aren't the models also distributed to various data centers--i think it's very easy with resources
101.
by Davidzheng 4mo ago
Human user usage data is probably a tiny contribution to improvement of the models--it's mostly RL on environments
102.
by Davidzheng 4mo ago
What is this comment? If they occurred we would face a huge disaster; isn't it better to err on the side of caution to make that risks as low as possible???
103.
by Davidzheng 4mo ago
I think it's too early to understand the ramifications but I agree this is a huge deal.
104.
by Davidzheng 4mo ago
Is the total compute capacity outside of meta, google, amazon, anthropic, oai and x is higher than even the capacity of any of them? In any case, there's no chance a public collaboration gets to anthropic levels of compute even if comm
105.
by Davidzheng 4mo ago
At some intelligence capability there can be catastrophic risk, the fact that we don't yet have any catastrophe doesn't mean the risk wasn't real. It's similar to new viruses which don't lead to outbreaks, the corre
106.
by Davidzheng 4mo ago
No...
107.
by Davidzheng 4mo ago
can I put open math problems on this site too with bounties
108.
by Davidzheng 5mo ago
someone correct if i'm wrong, but I think the max models are usually non-open
109.
by Davidzheng 5mo ago
My only gripe is that on my phone sometimes it takes like 30 seconds to load, which doesn't seem to happen for almost anything else
110.
by Davidzheng 5mo ago
Deep think still makes many many many more mistakes than gpt 5.5 pro on math
111.
by Davidzheng 5mo ago
I don't think there is a fundamental divide between implementation speedups and optimization and algorithmic/architecture optimizations
112.
by Davidzheng 6mo ago
But would miracle drugs and trading algorithms be as profitable as AI research/chip design/energy research? Probably if AI is by far the biggest growth in the economy majority of the AI's usage internally should (as incentivi
113.
by Davidzheng 6mo ago
There's still RL
114.
by Davidzheng 7mo ago
I agree it's not cheating that restricted sense. But I'm not really convinced that it can't be cheating in a more general sense. You can try like 10^10 variations of harnesses and select the one that performs best. And probab
115.
by Davidzheng 7mo ago
There's no objective measure of intelligence comparisons, we only say llm is jagged compared to humans.
116.
by Davidzheng 7mo ago
How can you tell?
117.
by Davidzheng 7mo ago
Important to remember that intelligence is not a singular thing and when the last gap is closed, most aspects will be highly superhuman
118.
by Davidzheng 8mo ago
Obviously yes in the form that the comment you replied to refers to--US would be much more careful stringing a country with nuclear weapons. So while the invasion may not be caused by proximity it can be allowed bc Iran doesn't have on
119.
by Davidzheng 8mo ago
How much value is there in individual values? Many of us remember that OpenAI was also started by people with strong personal values. Their charter said that they would not monetize after reaching AGI, their fiduciary duty is to humanity, a
120.
by Davidzheng 8mo ago
Election odds, chance of US bombing Iran, and many others

More ›