@SnapDragon's banner p

SnapDragon


				

				

				
1 follower   follows 4 users  
joined 2022 October 10 20:44:11 UTC

				

User ID: 1550

SnapDragon


				
				
				

				
1 follower   follows 4 users   joined 2022 October 10 20:44:11 UTC

					

No bio...


					

User ID: 1550

For the past decade at least, Apple has been the only tech company that operates on the principle of developing products and charging customers for them. A radical prospect, for sure, but in today's tech environment, not being dynamic seems like a good thing. Even their AI products seemed measured and promise to solve particular problems, not items that seem to be capable of everything and nothing at the same time.

As someone who was on Apple's AI/ML team until retiring last year, let me say that assigning intentionality to their AI failures is rather generous. The reality is that Apple's strengths have always been in hardware, not software. They were briefly the face of AI when they bought out the Siri team, but that didn't last too long. I'm kind of shocked Apple is doing so well in the market (...but grateful, given how overleveraged I am), since they're destined to be a consumer of AI, not a producer.

Yes, not all arguments are worthy of respect. And congratulations on pattern-matching a fallacy claim to an absurd extreme. If you think you've identified a systematic reason why all the smart experts in a field might be mistaken, cool. If you think that reason is that none of them know basic logic ... this is pure idiocy. You live on a planet filled with other smart humans. You are not the protagonist of the scientific world. Your design for a perpetual motion machine is wrong and your proof of Fermat's Last Theorem is flawed.

Let me tell you a story from when I was a snot-nosed, arrogant kid who had never met an intellectual equal at his age. I loved to read Scientific American. At one point I read an article talking about how how physicists were confused by how Black Holes seem to "destroy" information, while most physical laws preserve it. I decided then and there I wanted to mail Stephen Hawking and tell him that just because information is usually preserved, that doesn't mean it's a law of the Universe. I only had the pop-sci representation of an extremely complex topic. I didn't even know calculus yet. I barely understood what General Relativity was beyond the science-center demonstrations of a ball spiraling into a sloped hole. But somehow, from an article that was targeted above my age, I thought I had noticed that Prof. Hawking was making a simple logical mistake.

When I told my father this ... he, uh, suggested that I should not write that letter. It's funny to think about, in retrospect. I bet he wasn't quite sure WHAT to say to me. He wanted to preserve my feelings. But I have no interest in preserving yours.

This is the level of mistake YOU are making. Be better than my 10-year-old self. I would recommend, at an absolute minimum, actually reading the (extensive) Wikipedia page and seeing if you can find your purported logical error represented anywhere on it.

Sigh. I don't know why I'm wasting so much time on you being Wrong On The Internet. I do respect the hell out of your username, at least.

probably most coders in the game industry are using it, judging by noticeable sudden acceleration in output.

Do you have some good examples of this? I've been expecting to see it (and looking forward to it), but of the indie/AAA games I play, I wouldn't say they're coming out noticeably faster, more bug-free, or grander in scope. It's possible I'm just unobservant, or maybe (since dev cycles still exist) the real effect will still be delayed by a year or so.

The board does not have the right to light $1 trillion of shareholder wealth on fire so they can moralfag.

Right, that's why we have government!

That is not at all self-evident. I have seen no such increase in capability, and I still consider LLMs to be poorly suited for the programming world despite the continued hype.

Is your claim that there is not much difference between, say, GPT 6 and GPT 4o in coding capability? Do you reject all benchmarks and the revealed preference of programmers? I could understand (but not agree with) the skepticism a few years ago, but at this point you're clinging to a floating plank and loudly telling people who float by that there is no tsunami.

The argument sort of shifts though -- "sending probes to explore is normal, we've even done it" moves to "sending probes for conquest/destruction is normal, anyone would do it".

Well, no, the argument is "sending probes for conquest/destruction is normal, somebody might do it", and your counter-argument is "nobody would ever do it". Your claim is the much stronger one, and requires a stronger argument. This is the issue with a lot of simple-sounding "answers" to Fermi: they don't just have to reasonably apply to some or even most civilizations. If life is common in the galaxy, we need some sort of explanation as to why none of them chose to conquer the whole thing.

The Great Filter argument is that there's some trap in the future that wipes out almost all civilizations. The Dark Forest argument is that everyone has to keep quiet or they die. Given just how varied we think life can be, it's hard to find reasonable-sounding universal arguments. And, if none of them are true, we pretty much just have to conclude that we're alone (or very close to it) in the Milky Way.

You are not identifying a clever new insight that would make generations of cosmologists reconsider their careers. You are, and I want to make this point as clearly as possible, literally literally LITERALLY claiming that the Manhattan Project folks were unaware of the problem with generalizing from a sample size of 1. Tell me, do you regularly email Terence Tao to ask whether he's aware that sqrt(-1) doesn't actually exist?

Ok, um, while there are plenty of alien-life and even SETI proponents who are deeply silly and are just spouting pop-sci nonsense ... did you seriously suggest that the father of nuclear energy, Nobel Physicist Enrico Fermi, was among them? A lesson in intellectual humility: if you think you've solved a field in 5 minutes, you probably don't know what you're talking about.

Something you got completely wrong: if our civilization is somehow "normal", then a fellow civilization at the level of Renaissance Italy would be within 500 years of our level of development. On the timescales of our Universe, that would be an almost impossible-to-believe cosmic coincidence. Half the intelligent species in the galaxy should be millions of years more advanced than us, and we don't know of a clear reason why at least some of them wouldn't/couldn't have colonized a large proportion of the galaxy in that time. (Colonizing the galaxy at 0.1c is many many MANY orders of magnitude easier than "constructing Dyson Spheres".)

We know a lot of proposed answers to this so-called "paradox" - all much smarter than your casual dismissal - but there isn't one clear winner in the debate. We might still be wondering about this as we ourselves set out to the stars.

Also Western women range from 3 to 8.1 and Asian women range from 8 to 11. This is an undisputed fact and if you argue with me then you're wrong.

I generally agree with you that the industrial revolution was somewhat miraculous (and I appreciate all the historical detail!). I'm just not as certain that history couldn't have taken a slightly different path to the same endpoint - are you confident that each of your 8 steps are an inseparable part of the path to industrialization? Or, indeed, couldn't the same factors instead have aligned 3,000 years later, which would make zero difference to the Fermi Paradox?

There's a classic question of just how inevitable history is. There's Great man theory, which posits that it's highly dependent on the tiniest chaotic factors: the existence of particular individuals. On the other hand, a lot of innovations in math, science, and engineering were developed in multiple places simultaneously (calculus being one famous example), suggesting that technology might just progress in a somewhat straight line until certain advances "crystallize" out of it.

And at the timescales of the Drake Equation, you can't just argue that we were lucky to hit the industrial revolution quickly; you have to argue that, without X, we wouldn't have hit it in a million years (or however long it would take for a thriving medieval-level species to get wiped out by a global event). X = "cheap accessible hydrocarbons" does make for a reasonable hypothesis here.

Note that we don't use fossil fuels for "space travel" - the price point for that is well above the level where it's worth synthesizing better fuel. But I think your argument is fine if you just consider the industrial revolution instead, which was deeply dependent on cheap fossil fuels.

If we look at just humanity, society is 20k year old versus 100k years of behavioral modernity (or more!) so that's a rough 1/5 chance. If we extend to all of hominid history (~2 million years) now it looks more like 1/100. If we compare to the intelligent clades, with have maybe 2/6 clades (counting humans and LLMs against corvids, cetaceans, cephalopods, and elephants) and therefore maybe a 1/300 chance. Then compress to just the amount of time where we have space travel and we're looking at a 1/60,000 chance of getting a dominant intelligent species likely to develop space travel.

This is more of a quibble, but I don't think this is the right way to look at things at all. Suppose, hypothetically, that all intelligent species were guaranteed to produce a spacefaring civilization after some time, following the same exponential rate of improvement as us. Then for any level of technology you could always look back and say "gosh, it took us [X years] to get here and we've only had [technology] for [Y << X years] - the chances we would develop [technology] were minuscule!" When, in fact, the actual probability is 100%.

You really do just have to Do The Work and analyze all the specific factors you brought up. Intelligence level, temperament, access to resources, climate, lifespan, population size - in how many of these respects did we indeed "get lucky"? Also, did we manage to soldier on through despite certain disadvantages? What are the limits of the resourcefulness of an intelligent species determined to claw its way up the tech ladder? e.g. imagine some other civilization thinking "no way could we have reached the space age without our 1,000 year lifespans - trying to condense learning into 30-40 years would make scientific progress impossible!"

We "outran" evolution by several orders of magnitude, so I don't think building a technological civilization is a "race against time" in any real sense. Also, most of our civilizational development could have happened without hydrocarbons. Coal was definitely valuable to early civilizations, but charcoal would have sufficed, pre-industrial-revolution.

It's a little hard to point at specific human advantages and say they were essential for us to reach the modern age. Sure, our modern civilization was built almost entirely around hydrocarbons, but that only shows that they were the cheapest option we had, not that they were the only option we had. There's the classic question of whether superintelligent dolphins could ever conceivably make a technological civilization. My guess would be no, since they lack hands necessary for developing tools, easily accessible metals, and fire ... but who knows what alternatives 100,000 years of civilizational experimentation could come up with? It's difficult to prove a negative.

But this in turn debunks the great filter arguments people(mainly liberals) make about all sorts of pet issues- the great filter is already behind us.

Some rationalists also point to the Doomsday Argument, which says we're all going to die for a completely different reason. I just shake my head in despair whenever Scott brings it up. The argument is completely wrong, as I try to point out every so often. But I'm fighting against Bostrom, against a Wikipedia article guarded by proponents, and against a population of people who enjoy thinking they've learned a secret hidden truth.

Well, the Drake Equation was an attempt to be a little more formal with that logical leap. We do have observations or reasonable hypotheses about several of the important factors. To counter-analogize, imagine a universe filled with a soup of DNA and some intelligent life, that notices its own DNA is around 2000 symbols long and can estimate the rate of variance in the DNA soup. These beings could put a decent lower bound on the probability of other intelligent life existing, even with no other examples.

In practice, though, I do tend to agree with you that there's not much reason to be "surprised" that we seem to be alone in our galaxy. As you said, for anthropic reasons, the observation "we exist" does not give us much information on its own. The Universe doesn't owe us comprehensibility, and some of the factors in the Drake equation may just be inestimable, despite our best efforts. Which makes the whole thing not much better than a shrug.

I don't know. The population of rabid antisemites who are still cogent enough to hold their own in a Motte discussion is probably quite small - but we probably ended up with ALL of them. It's the classic witch effect.

Oh, does The Motte have a warrant canary? This is news to me.

It helps that I have literally nowhere else to go. I'm not aware of a single other forum where both sides of trans issues can be discussed. As for AI, I occasionally try to make anti-doomer comments on ACT, but Scott and most of his community are true believers in the Cult of Yud and I typically get mobbed. I'm guessing LW would be even worse.

It just seems like first world problems or a skill issue to complain too much about the modern world in any way (I'm only talking about first world western nations for the entirety of this conversation). It's a really nice world we've been given. Just the right difficulty setting, so to speak.

While I don't share your joie de vivre, you are objectively right about this, and it needs to be constantly shouted from the rooftops. Relative to all humans who've lived before, we've carved out an amazing paradise from the brutal sandbox that is the Universe.

I'm the same, though I haven't had a near-death experience in a while so it's hard to judge how I'd genuinely react. It does make it weird when I'm watching a movie (especially horror movies, which I quite enjoy) and I see somebody going to extreme lengths to stay alive. I just have this disconnect where I'm thinking "why would you bother?". When I talk to other people about death, they are quite scared of non-existence, and it's hard for me to relate. The crappy part about dying is the pain of dying, and how your friends/family will miss you. But non-existence? You were non-existing for billions of years before, and it didn't seem to bother you.

Well, at least you can take comfort in having been right (even if it was kind of obvious if you were paying attention). But this does seem to be one of the many dependable aspects of society that will be mostly gone in the AI future - you really won't gain much relative status from being right/talented/hard working. The playing field will be leveled for everyone, which means no incentive to truly excel, except for personal satisfaction. (EDIT: Assuming that smart models remain accessible to the masses, rather than hoarded by elites.)

Huh, I actually hadn't noticed that; I did enjoy (most of) the characters. Was it really so blatant?

Well, this is why it's good to get a second opinion. Thanks. I guess all this culture warring really has made me too sensitive.

I'm pretty economically ignorant myself, so I guess that side of things didn't bother me as much. But I do have a related criticism that a barely-industrialized world clashes with modern libertine sexual and gender politics. Magic or no magic, it makes no sense to pretend that gay relationships and casual sex are "just normal" in a society that's barely above subsistence level. I suspect that's just what Royal Road readers are used to: shipping characters they like with no respect to realism or gender, and naive anti-capitalist ideals.

So I guess I agree that the worldbuilding just doesn't hold up under even casual scrutiny. Still, I do appreciate the rational way the MC goes about trying to figure things out in the time loops. It's got the nice quality that the plot is mainly driven by lack of information, not bad decisions (or at least, not unrealistically bad ones).

Though I'm liking Throne of TIme a lot so far!

I'll have to check it out!

In some cases the observations are actually correct (yes, all the women are romantic interests and are not the action heroes the men are), but that is not necessarily a bad (or unrealistic) thing unless, well, you're a feminist reader who demands all stories give "equal representation" no matter how it distorts the story.

I worry that all this discourse has broken my brain, because I've found myself doing the opposite in a current series I'm reading. It's called Years of the Apocalypse. It's a time-loop webfic, and the best I've read since Mother of Learning. But I just couldn't help but notice that the author wanted you to treat a character sympathetically or not based on their gender. The main character and 95% of the people moving the plot forward are female, with no good reason for it. It's really bad if you analyze the irredeemable villains; out of dozens, it took until book two before there was a SINGLE female one.

I wish I could just ignore it, because if I were to suddenly become unable to distinguish pronouns, it'd become a book filled with varied, believable, and interesting characters. But I've become tuned to Notice these things, largely by feminists complaining about the converse. It hurts the story either way.

After all, the Big Lie so many feminists told was that they just wanted equality. That's why "feminism" has such positive connotations in most people's minds, despite literally being an -ism. But the moral people who believed in equality have largely gone their own way; the feminists who are left over actually just think that sexism is only wrong when it's done TO them, not BY them.

Huh... what don't you like about the current remake trilogy? I admit it's still on my list to play, but it certainly seems like they've put a massive amount of love and nostalgia into it.

Same here. I do think we're on the fast track to gaming experiences that truly would have seemed like ridiculous sci-fi ten years ago. Not quite there yet, of course, but there are signs. I consider it a significant milestone that there are now tons of bespoke AI music videos on Youtube (and, I'm sure it's just my lack of taste and discernment, but I find a lot of them very very good). Going from niche personalized music videos to niche personalized movies to niche personalized videogames doesn't seem like it's going to take too long (though it's possible we'll hit a compute wall somewhere).

You know what would truly make my Glorious AI Future? AI creating a Freespace 3 that satisfyingly finishes the story.