eru

8 ms·

181.
by eru 1mo ago
There's lots of competition in AI. Where do you see the 'insurmountable advantage in everything'?
182.
by eru 1mo ago
The trailing edge of AI is catching up fast. There's plenty of open source (and even more open weight) AI models and they are getting better and better. If that's your only objection: in a few years you can prove Rieman's hy
183.
by eru 1mo ago
> Does mathematics still have a practical impact without humans in the loop? There's a single answer to that question: yes, math very much has an impact without humans in the loop. Math has a lot of applications, and those applicati
184.
by eru 1mo ago
You say it like it's a bad thing.
185.
by eru 1mo ago
Well, what you are saying that the situation is even better for math than for chess? Chess is only valuable as an entertainment. So no one really gains anything from computers becoming really good at chess. But with math, everyone in the w
186.
by eru 1mo ago
Who cares about the giant labs? The trailing edge will catch up quickly, and in a year or three you can prove Rieman's hypothesis on your smartphone.
187.
by eru 1mo ago
> the models capable of this work are not always going to be exclusively under the control of those aligned parties. Yes, the trailing edge will catch up quickly. I wonder how long until P vs NP falls.
188.
by eru 1mo ago
You paid for exactly what the license agreement says. And that includes that Sony can pull the plug. If you still pay, that's on you.
189.
by eru 1mo ago
If you don't like the burger at McDonald's, you go across the street to Burger King. Lots of power for the customer, even though the offer is take it or leave it.
190.
by eru 1mo ago
For the US, you really need to be careful to get something that doesn't fall out of the plug. For the British standard, pretty much any plug you can buy will be fine.
191.
by eru 1mo ago
US plugs (and similar ones like Japanese and to a lesser extent the Australian ones) are really bad. European ones mostly stay in the socket. But I guess they ain't quite as cleverly designed as the British ones.
192.
by eru 1mo ago
You give workers more power by making sure there's plenty of competition between employers, and plenty of other outside options for workers. See eg how here in Singapore we have no minimum wage and basically no unions worth that name,
193.
by eru 1mo ago
British food before roughly WW2 wasn't so bad actually. The French have some decent bread, if you can live without the taste of rye for a while.
194.
by eru 1mo ago
In Germany your employment contract can have Vertragsstrafen, ie the contract can say that if you do something you need to pay a penalty even above and beyond damages. But in any case, employment contracts are weirdly restricted in many pla
195.
by eru 1mo ago
Are you being sarcastic?
196.
by eru 1mo ago
The plug ain't too bad actually. (And I grew up with European style plugs.) Still a vastly better mechanical design than what they make you use in the US: they barely hold in the socket.
197.
by eru 1mo ago
Alas, Chorleywood tastes awful. At least to German taste buds.
198.
by eru 1mo ago
The solution posted still requires O(n) calls to your random number generator. You can do it in something like O(log n) calls while still sticking to a single forward only pass. (But compared to reading the filesystem the rng calls were pro
199.
by eru 1mo ago
Being able to restrict yourself is exactly what the right to enter contracts is. Just like being able to commit yourself to showing up to work everyday restricts your freedom. But it's essential to getting the other party to give up s
200.
by eru 1mo ago
See https://sockpuppet.org/blog/2015/03/06/the-hiring-post/ for some ideas.
201.
by eru 1mo ago
Depends on what you do in life.
202.
by eru 1mo ago
You can play a variant of math where you only have a limited set of options available each turn, and it wouldn't make much of a difference. (Eg limit yourself to commands you can send to vim as you write your Lean proof or whatever.)
203.
by eru 1mo ago
> I think in general, any game that is 'turn-based' (in this case chess), with a limited set of options, quickly boils down to pure theory and in my opinion 'dumb intelligence' of just remembering many things and patt
204.
by eru 1mo ago
You can play Diplomacy for that.
205.
by eru 1mo ago
Your theory of how companies work is certainly interesting. It sounds like you think they have solved the principal–agent problem? https://en.wikipedia.org/wiki/Principal%E2%80%93agent_proble...
206.
by eru 1mo ago
I don't see the cascading hierarchy of enrichment. I mean, they are certainly trying, but so far there's too much competition so the surplus mostly goes to customers.
207.
by eru 1mo ago
> 1M cached tokens on deepseek is $0.006, the big labs can't sell anywhere close to this, they have funders expecting returns and huge overhead. Losing most of your customers tends to sharpen the mind a bit. They could eg stop push
208.
by eru 1mo ago
> But the counter to that is that everyone else is becoming more compute efficient so maybe that advantage isn't what people thought it would be. I'm not sure? If we have techniques to use the hardware even better, that will m
209.
by eru 1mo ago
You could also have eg Sol do the babysitting.
210.
by eru 1mo ago
I did a lot of that kind of work with the older version of DeepSeek before they upped the prices. For example, it was quite good to get a decent Sashiko review. Sashiko is a Linux kernel review agent with interchangeable LLM driver. It&#x

More ›