myrmidon

16 ms·

331.
by myrmidon 9mo ago
Your article is careful to never explicitly state correlation between the vaccine and those seven girls deaths. Without such a link, your argument falls apart. > So answer me honestly: Did that same billionaire (Bill Gates) and his orga
332.
by myrmidon 9mo ago
> The crossroads where one path leads to rule by entrenched and corrupt oligarchs, at least as bad as any of the court eunuchs of old, and where the other path leads to ochlocracy. I'm a bit confused; assuming you are aiming at the
333.
by myrmidon 9mo ago
You state "Bots that talk to each other" as discovery, but on the security page you describe the involved accounts as spam ring: How is that not completely invalidating?
334.
by myrmidon 9mo ago
I do agree with this. If you followed this approach consistently, you would need back pressure against individual and company wealth growth. This could be quite good for competition, but would probably hurt sectors a lot that have high fixe
335.
by myrmidon 9mo ago
I'd expect to find most of the incriminating stuff in emails and short messages. > And internal comms could be asked for without the nuclear option of raiding the office Relying on the accused to compile and provide evidence seems n
336.
by myrmidon 9mo ago
> I don't know about that. Would "I didn't know it was real" really count as a legal defense? Absolutely-- prosecution would presumably need to at least show that you could have known the material was "genuine&
337.
by myrmidon 9mo ago
> Saying 'this makes enforcement of other laws harder' does not do that. You could use the same reasoning against encryption. Yes. I almost completely agree with your outlook, but I think that many of our laws trade such indivi
338.
by myrmidon 9mo ago
If there is a glut of legal, AI generated CSAM material then this provides a lot of deniability for criminal creators/spreaders that cause genuine harm, and reduces "vigilance" of prosecutors, too ("it's probably ju
339.
by myrmidon 9mo ago
There are multiple valid reasons to fight realistic computer-generated CSAM content. Uncontrolled profileration of AI-CSAM makes detection of "genuine" data much harder, prosecution of perpetrators more difficult and specifically
340.
by myrmidon 9mo ago
In such a case specifically: Uncover internal communication that shows the company was aware of the problem and ignored it, which presumably affects liability a lot.
341.
by myrmidon 9mo ago
> Nobody can publicly admit that it's fine to letting a 5-year-old kid die once in a while, even if that would be the right call. Sure, there are such cases, but a lot of regulation was written in blood, and the price that affected
342.
by myrmidon 9mo ago
Sure, but how many "correct" decisions are not made or drag on forever because of vetos? Allowing veto power to single participants is often crippling for institutions in practice, because you allow every political adversary (
343.
by myrmidon 9mo ago
What? No? The main selling point is eliminating costs for a human driver (by enabling people to safely do other things from their car, like answering emails or doomscrolling, or via robotaxis).
344.
by myrmidon 9mo ago
Humans can also be just a tool, and have been successfully used as such in the past and present. But I don't think any slave owner would sleep easy, knowing that their slaves have more access to knowledge/education than they thems
345.
by myrmidon 9mo ago
If we had human level cognitive capabilities in a box (I'm assuming we will get there in some way this century), are you confident that such a construct will be kept sufficiently isolated and locked down? I honestly think that this is
346.
by myrmidon 9mo ago
Relevant excerpt from your own wiki guideline: "Do not rely too much on your own judgment. [...] if you are an expert user of LLMs and you tag 10 pages as being AI-generated, you've probably falsely accused one editor." Never
347.
by myrmidon 9mo ago
> Who cares who owns the data center? The govt can send in the army and nationalise it Do you think the government is going to react with mass nationalisation of private companies to fight wealth inequality? What would the threshold be?
348.
by myrmidon 9mo ago
"Disruptive rethinking of economics" is a very optimistic way to put this IMO. The big problem I see is that there is little incentive for "owners" (of datacenters/factories/etc) to share anything with such h
349.
by myrmidon 9mo ago
Any form of AI unconcerned about its own continued survival would be just be selected against. Evolutionary principles/selection pressure applies just the same to artificial life, and it seems pretty reasonable to assume that drive&#x2
350.
by myrmidon 9mo ago
A big short-term risk that I see is that AI is going to cripple wealth redistribution mechanisms that we currently rely on. Most willing persons have access to income by providing labor right now. If the value of that labor diminishes becau
351.
by myrmidon 9mo ago
Not so! Concentration camps predate the Nazis by quite a bit, and the term was/is used for all kinds of prison-like setups where inmates are held outside the rule of law. E.g. US camps holding Japanese immigrants during WW2. Sure, it m
352.
by myrmidon 9mo ago
Having the government manipulate mass-media material is frankly insane. Everyone involved in something like this should be sacked immediately, but I have very little hope that voters are going to punish egregious misbehavior like this as lo
353.
by myrmidon 9mo ago
I think this is a scarily prevalent view, and I strongly disagree. Regarding defense: The US is not the world policeman and never really has been. Police enforces higher laws-- but the only law the US enforces is its own, which sometimes
354.
by myrmidon 9mo ago
Completeley agree. It's such a stark contrast between Bush senior and junior, too-- GWB put the US on such a bad trajectory (middle east interventionism, surveillance state, etc.), altough 9/11 probably deserves a lot of blame f
355.
by myrmidon 9mo ago
Because it's reductive bullshit. Energy price is just one of many inputs for the viability of industry. Availability of (educated) labor, wage level, infrastructure, political stability and a ton of other factors are at least as if not
356.
by myrmidon 9mo ago
Maybe the CRC32 implementations in the smasher suite just aren't that fast? [1] claims 15 GB/s for the slowest implementation (Chromium) they compared (all vectorized). > The 32 bit hash of CRC32 is too low for file checksums.
357.
by myrmidon 9mo ago
> I'm pretty sure xxhash is a straight upgrade compared to CRC32. Unclear; performance should be pretty similar to CRC32 (depending on implementation), and since integrity checking can basically be done at RAM read speeds this shoul
358.
by myrmidon 9mo ago
We do know that they are different, and that there are some systematic shortcomings in LLMs for now (e.g. no mechanism for online learning). But we have no idea how many "essential" differences there are (if any!). Dismissing LLMs
359.
by myrmidon 9mo ago
> If you had enough paper and ink and the patience to go through it, you could take all the training data and manually step through and train the same model. But you could make the exact same argument for a human mind? (could just simula
360.
by myrmidon 9mo ago
This is exactly why its so hard to compare though; government contracts, emission credits and direct subsidies are all quite different and weighting them is highly subjective.

More ›