theptip

9 ms·

151.
by theptip 4mo ago
The shallow answer here is: AI is already being asked to simulate human-like agents with self-preservation. Of course more realistic simulators will be put to this purpose too! And by evolutionary pressure, the ones with self-preservation w
152.
by theptip 4mo ago
Right. And if one actor happens to accrue greater power you would have less disincentive to crush your competitors. (Honestly this could happen even before people-less economies, but the disconnect from democratic opinion is even stronger w
153.
by theptip 4mo ago
The important question is, who has power in such an economy? In the current economy by necessity labor and capital are both required, and when capital tries to subjugate labor there tends to eventually be a violent reaction. Given the depen
154.
by theptip 4mo ago
Thanks for engaging with this in a quantitative way. I particularly appreciate the earnings/valuation accounting, this really helps anchor the discussion. I agree you’ve demonstrated that there is enough economic growth to not contradi
155.
by theptip 4mo ago
And yet the N=1 he chose doesn’t disprove the statement.
156.
by theptip 4mo ago
It’s a good point, I should have compounded the 15% m/m to compare with 2.5% economy annual growth rate.
157.
by theptip 4mo ago
Fwiw I agree that the universal impossibility statement is too strong. But his example doesn’t demonstrate anyone earning $1b. It just demonstrates a very high growth rate at $1-2m.
158.
by theptip 4mo ago
This is the level of detail I am asking for. “Subpopulation height increases because of the physical and understood processes of maturation and growth”. You could easily go into the biology involved, model the genetic, environmental, and ra
159.
by theptip 4mo ago
How do you grow better than average?
160.
by theptip 4mo ago
I’m in favor of this, but it doesn’t solve the full problem. If all your friends use social media as the fabric of their social interactions, you’ll be ostracized if you opt out as an individual. IOW its a coordination problem. You need mos
161.
by theptip 4mo ago
Counterpoint - this is a coordination problem, there are studies suggesting that most kids would rather not participate in the whole social media thing but an individual can’t opt out. It’s very possible that a policy like this could give e
162.
by theptip 4mo ago
But his N=1 anecdote doesn’t prove anything. He shares a feel-good story about an early stage company with very high growth on a small base. This person is not a billionaire yet. The actual comparison would be to look at all the startups wi
163.
by theptip 4mo ago
Right, regardless of whether you agree or disagree with the point, PG doesn’t actually engage with it. He just says “compound growth + build something that people love”. But the meat of the point is: if the economy is growing at 2.5%, how d
164.
by theptip 4mo ago
So you agree this is not what Anthropic wanted?
165.
by theptip 4mo ago
> The only way to prevent that one entity weaponizes it, is by giving EVERYONE access to it There is a middle way; the policy space also includes government regulating both access and monopoly. I’m opposed to monopolies of this tech, but
166.
by theptip 4mo ago
This is nonsense. What Anthropic have been campaigning for, since the beginning, is a principled rule-based audit of model releases. Now we are getting reactive, arbitrary and capricious enforcement; rules rushed out the door; classified ev
167.
by theptip 4mo ago
> I have no domain expertise that another Sr. engineer steering an LLM cannot match. All my finance and payment domain expertise, all the debugging intuition and distributed system knowledge earned through hours of sweat and tears, is no
168.
by theptip 4mo ago
There is a sense in which it doesn’t matter at all; many of the limitations of agents in large codebases are just the context management challenges. So proving that you can cohere and progress at O(1m) is a useful scale observation. “Can I
169.
by theptip 4mo ago
For sure. Social engineering attacks on human support staff are common and well known, but the skill floor is non-trivial; you need to actually be able to convince a human of your ruse. Having a support agent likely made it easier to enumer
170.
by theptip 4mo ago
Sounds like they are saying the agent did not malfunction, and this vuln could have been triggered by a human support agent too.
171.
by theptip 4mo ago
Nah. Skills are great. But you should write your own.
172.
by theptip 4mo ago
I don’t have a horse in the race, but the OP objections ring hollow to me. - scope is important: true, but isn’t that derivable from the commit contents? An important sanity check on a diff is to look at the paths touched. (A “test” diff sh
173.
by theptip 4mo ago
Maybe more people if the contract is that it’s how you get your work into the release notes?
174.
by theptip 4mo ago
If you’ve got a k8s cluster, review apps are completely solved, right? You only need this if you’re in the niche of “no k8s, no PaaS”.
175.
by theptip 5mo ago
Right, perhaps true, but my point is that you don’t need a consumer economy if/when you have fully replaced labor with automata. I agree this would represent the end of the current socioeconomic system, and that many/most would no
176.
by theptip 5mo ago
No.. that’s not what AI succession means. This all supposes a powerful and capable enough AI entity that there is no human in control. Rather than simply asserting that your interlocutors are wrong, you are welcome to advance an argument fo
177.
by theptip 5mo ago
I agree with all these observations. This is the best argument for successionism IMO. If you can be confident that you are creating a BDFL that is genuinely better than human leaders (a quite low bar) then it seems a good trade, unless you
178.
by theptip 5mo ago
This is only a bargaining chip until the robot workers come online. Once capital can fully replace labor it’s game over. (And once the robots outpower the humans, you can’t even protest or overthrow your new overlords.) Important to know wh
179.
by theptip 5mo ago
> trying to preserve the human species as it is would be silly There are a few related concepts being conflated here. The quote above is a Transhumanist position; they think that humans need to evolve. I don’t think most transhumanists a
180.
by theptip 5mo ago
I think this neatly captures the gap in the article; it’s thinking of domain expertise as static, rather than something that needs to be built/extracted. It also goes the other way; A a good engineer can build experiments and discover

More ›