bl0rg
bl0rg on Hacker News
Well if it works on larger networks, including itself, it's certainly a much faster version of recursive self-improvement than proposing a new architecture and spending 500 million dollars on GPU training.
At some point someone will analyze this pattern and post an article named "The Unreasonable effectiveness of the 'The Unreasonable X of Y' template".
> Which is knowledge that really should factor into the decision to build the next, bigger one.
It was always factored in, and of course it would be in any next iteration.
> Invest the money and effort elsewhere, for now. There are many other fields of scientific exploration that are very likely to yield greater return (in knowledge and utility) for less. You could fund a hundred smaller but still substantial intiatives instead of one big accelerator. And be virtually guaranteed to have an exciting breakthrough in a few of them.
I agree with this to a large extent. I'm just not against particle accelerators as a venue for scientific advancement and in the best of worlds we could do both.
No, I'm not and neither is anyone else. It's common sense that we should explore options that require less effort, just as one would in any project. I'm saying that we can't discredit huge particle accelerators due to, in the grandest scheme of things, a small economic cost and past results of a different experiment.
> The latest ones cost billions and have confirmed a few things we already thought to be true.
Yes, but we had hopes that it would lead to more. And had lead to more, something only known to be false in hindsight, who knows where that would have ended us up? What if it upended the standard model instead of reinforcing it?
> Absolutely not.
What are we supposed to do then? As humans, I mean. No one knows why we're here, what the universe really is like. We have some pretty good models that we know are wrong and we don't know what wonders the theoretical implications of any successor models might bring. That said, do we really need to motivate fundamental research into the nature of reality with a promise of technology?
I'm not arguing for mindlessly building bigger accelerators, and I don't think anyone is - there has to exist a solid line of reasoning to warrant the effort. And we might find that there are smarter ways of getting there for less effort - great! But if there isn't, discrediting the venue of particle accelerators due to their high upfront cost as well as historical results would be a mistake. We can afford it, and we don't know the future.
I'd rather kiss my dog, whom I love, over a random person that I have little in common with. Is that somehow weird to you?
No, that's not what I'm implying. I'm saying that it's needlessly complicated.
I guess it depends what we mean with straightforward. If we mean something along the lines of "no ambiguity" then yes. If we mean something along the lines of "simple, easy to do" then no. Almost anything can be accomplished with a sufficiently long checklist. I just feel like the entire process could be streamlined and simplified.
Starting a company in Sweden requires (uploaded PDF from Bolagsverket to ChatGPT who summarized):
1. Prepare the foundation deed and the articles of association.
2. Identify the beneficial owner(s).
3. Pay the share capital and obtain the bank certificate or auditor’s statement.
4. Submit the registration application for the limited company to the Swedish Companies Registration Office (Bolagsverket) and wait for approval.
5. If applicable: submit a certified copy of your passport (non-Swedish citizens).
6. Apply for F-tax approval and VAT registration and wait for the decision.
7. Register as an employer if you will pay salaries.
8. Keep continuous bookkeeping and prepare the annual accounts each financial year.
9. Submit the annual report to Bolagsverket every year.
Optional:
1. Obtain business and personal insurance.
2. Register trademarks or protect other intellectual property.
3. Choose an auditor if you want one or when the company later reaches the required thresholds.
4. Register a cash register if you accept cash or card payments.
5. Meet requirements for import/export and obtain an EORI number.
6. Follow rules for buying/selling goods or services within or outside the EU.
7. Keep a staff ledger if required for your industry.
8. Follow reverse-charge VAT rules if you operate in construction.
9. Apply for permits if your specific business activity requires them.
This is not what I'd call a straightforward process, personally. Also speaking from personal experience. Sorry for the formatting.
I see no contradiction here.
It scares me that people think like this. Not only with respect to AI but in general, when it comes to other life forms, people seem to prefer to err on the side of convenience. The fact that cows could be experiencing something very similar to ourselves should send shivers down our spine. The same argument goes for future AGI.
This doesn’t sound reasonable to me. Why would we?
Why do you feel that way?
In a few years, will we need seniors?
I've tested it and it's actually quite good. I'll be switching.
Yes - only the Irish are brave.
What did you try to use it for? But I agree, it feels kinda beta at the moment.
Coincidentally that's also the framerate of the YouTube stream main camera (please fix OP)
A consciousness without memory (if that's even a possibility) would never notice any change.
It is and it isn't. I've gotten so used to the promise of AI that I don't know if I'll ever be able to go back to imagining a "normal" future again. Of course it can all turn to fuck, but it's so much more interesting than dying at 85 in some home.
Volvo is designed/engineered in Sweden and has a global production network with factories in Sweden, Belgium, the US, and China. You're massively oversimplifying.
As someone who has played both, why not just play Black Mesa?
It was a dumb attempt at a joke
Reality doesn't necessarily agree with its model though, so it makes sense to verify?
Can you explain what it means to reason about something? Since you are so confident I'm guessing you'll find it easy to come up with a non-contrived definition that'll clearly include humans and future "actual AI" but exclude LLMs.
Graphite is a pretty common word, and it's easy to make the connection between that and what graphene is. Same goes for gold and goldene. Not so much for chrysos and chrysene.
Running it for larger and larger numbers, will report back soon...