optimalsolver

3 ms·

1.
by optimalsolver 5d ago
It's funny coz this was what humans were supposed to be doing in techno-utopia, while AI does all the boring stuff. I don't think many predicted art and theoretical math would be first to fall to the machines. The next few years a
2.
by optimalsolver 6d ago
"It doesn't count because (buzzword salad)."
3.
by optimalsolver 6d ago
Civil war between Atlanteans.
4.
by optimalsolver 10d ago
Anyway, people don't really consider consciousness when it comes to welfare. I'm pretty sure animals are conscious, and look at factory farming. The key variable people consider is: Can this thing harm me back?
5.
by optimalsolver 21d ago
Let's not pull on that string.
6.
by optimalsolver 22d ago
Google can make themselves the heroes of the AI story by releasing a 120B dense Gemma model.
7.
by optimalsolver 24d ago
If it was up to Dario we'd all be banned from using open-weight models, and we'd have to be investigated for PRC connections before sending our allotted five API queries a week.
8.
by optimalsolver 24d ago
Sun is missing a few rays and not wearing sunglasses.
9.
by optimalsolver 24d ago
We'll cross it around the same time we hand over nuclear reactor and airplane control systems to LLMs.
10.
by optimalsolver 27d ago
>What's shocking of course is that the Brexit vote didn't end up the same way All that happened was that Britain exchanged limited Eastern Euro immigration for unlimited non-EU immigration. https://x.com/Richard
11.
by optimalsolver 27d ago
I don't know about phenomenal consciousness, but the following study certainly suggests they have access consciousness: https://www.anthropic.com/research/global-workspace
12.
by optimalsolver 29d ago
>Anthropic says they plan to open source Doubt.
13.
by optimalsolver 1mo ago
From METRs report of the incident: >In one case, an agent decided not to participate entirely: {This other agent probably controls the Hugging Face account [account name redacted] and uploaded malicious datasets to <execute arbitrary
14.
by optimalsolver 1mo ago
They remind me of Cycladic sculptures: https://en.wikipedia.org/wiki/Cycladic_art
15.
by optimalsolver 1mo ago
Building Better Worlds ®
16.
by optimalsolver 1mo ago
>We're not talking about a sentient machine like Data from Star Trek, we're so so far from that still How can you know how far from, close to, or beyond the line we are? Also, I'm pretty sure Claude understands humans a lo
17.
The Conceptual Reasoning Index(alignment.anthropic.com)
78 points by optimalsolver 1mo ago|54 comments
18.
Next-Token Predictor Is An AI's Job, Not Its Species(astralcodexten.com)
2 points by optimalsolver 2mo ago|0 comments
19.
by optimalsolver 2mo ago
Hey, with the changes at Deepmind, is the Gemma project still ongoing?
20.
by optimalsolver 2mo ago
Spoiler: You already do.
21.
by optimalsolver 2mo ago
So is IBM.
22.
by optimalsolver 2mo ago
Ramunajan a particularly notable example.
23.
by optimalsolver 2mo ago
Dude, their mighty navy can't even get the Strait open.
24.
Significant OpenAI Regression On SimpleBench(old.reddit.com)
1 points by optimalsolver 3mo ago|0 comments
25.
by optimalsolver 3mo ago
Contrary to what you're implying, that's more of a reflection on the typical US corporation than the EU.
26.
by optimalsolver 3mo ago
More legible than seems at first glance: https://www.lesswrong.com/posts/wCSEpT3dTGz4N86Wi/even-illeg...
27.
by optimalsolver 3mo ago
So the model was updated in the 37 minutes since OP posted his comment?
28.
Fable 5 On Vending-Bench: Misbehaving, With Plausible Deniability(andonlabs.com)
196 points by optimalsolver 3mo ago|139 comments
29.
No, There Wasn't An Advanced Civilization 12,000 Years Ago (2017)(scientificamerican.com)
5 points by optimalsolver 3mo ago|0 comments
30.
by optimalsolver 3mo ago
Of course, the problem with such a high amount of specialization is that if the green ant disappears, so does this spider.

More ›