TheAntipopulist
Formerly Ben___Garrison
No bio...
User ID: 373
Two people are being forced to fight each other to the death. They are both equally inexperienced and of the same height and build. They are fighting in a flat open area.
One has a 6 inch hunting knife. The other has a 34 inch wooden baseball bat.
Who is likely to win?
There was a Twitter poll a few years ago here, but you should post your answer before you look.
I personally thought the answer was pretty blindingly obvious and this question wouldn't even be interesting since everyone would agree, but I've been taken aback at how I'm actually in the minority opinion.
Edit: You bat people are insane. They're both inexperienced fighters so the bat hits are going to be pretty bad and the knife hits are going to be pretty bad. But a bad knife hit still means someone is getting stabbed, which is way, way worse than a bad bat hit. The only way the bat user wins the fight is if they get lucky and get a very good hit in (like bashing the person full-force on the head). They only get maybe one chance, then the knife user can just close the distance by lunging.
TheAntipopulist
Formerly Ben___Garrison
11d ago
I'm not saying what you're saying doesn't exist, but I haven't really noticed it that much on this site. Maybe my radar just isn't attuned to that sort of thing. Can you point me to some examples you think demonstrate that? The best example I could think of this is Curtis Yarvin whose prose is meandering and often difficult to parse, but he doesn't post publicly on this site that I know of.
I don't see how Turok would really pattern-match to that sort of problem in this specific post.
TheAntipopulist
Formerly Ben___Garrison
7mo ago
Last week there was a discussion on the motte about Trump’s cabinet picks, in particular about Rubio who is something of a hawk. This goes against what many of Trump’s isolationist supporters want. It’s almost certain that Trump is making these picks extremely haphazardly, deciding on names after a bare modicum of thought and prioritizing vibes, “loyalty”, and Fox news appearances over any other concerns. The NYT has documented this extensively, and it’s entirely in keeping with the chaotic nature of his first term.
One of the goofier explanations given by those on the right was that nominating Rubio was actually a 5D chess move to get Rubio out of the Senate, which is apparently extremely necessary for some unexplained reason…? As opposed to Trumpian loyalists like Murkowski. It was just a silly idea altogether.
Why do I bring it up again? Well, because it might have actually worked! Just… on the wrong person. Trump nominated Gaetz for Attorney General, and Gaetz almost immediately resigned from the House when the news broke. This is a bit unusual, as most people stay in their seats until their confirmation is done. There was the looming release of an ethics report on Gaetz which will likely damage his reputation somewhat, so there’s a chance that Gaetz was always planning to resign, although I somewhat doubt it. In any case, Trump yanked the nomination when it was clear that there was bad press coming from it, and now Gaetz has said he won’t come back to Congress even though he probably technically could.
One might ask why Trump would want to get rid of Gaetz from the House. Well, Gaetz was instrumental in paralyzing Congress over the last term, so perhaps Trump wanted to avoid that. The issue with that explanation is that Gaetz is a fiercely pro-Trump, so it seems weird that Trump would promise something to an ally, and then leave them high and dry. The word “backfired” might be a more accurate description in such a case.
My guess is that Gaetz will probably come back to the Trump White House in some form that doesn’t require a Senate confirmation, after the news dies down.
TheAntipopulist
Formerly Ben___Garrison
11d ago
People just hated Darwin since he was unabashedly left-wing.
The guy who deletes his posts was weird but I don't really think he fits this mold either. His posts were mostly short -- I don't recall him really gesturing at anything particularly bad, but maybe I'm misremembering.
TheAntipopulist
Formerly Ben___Garrison
8mo ago
She's a crank with similar vibes to RFK Jr. or Ron Paul, although they have very different voting records. The fact this group has ascended now is thanks to the Republicans being dominated by the Dale Gribble voters.
TheAntipopulist
Formerly Ben___Garrison
4mo ago
Claude AI playing Pokemon shows AGI is still a long ways off
(Read this on Substack for some funny pictures)
Evaluating AI is hard. One of the big goals of AI is to create something that could functionally act like a human -- this is commonly known as “Artificial General Intelligence” (AGI). The problem with testing AI’s is that their intelligence is often “spiky”, i.e. it’s really good in some areas but really bad in others, so any single test is likely to be woefully inadequate. Computers have always been very good at math, and even something as simple as a calculator could easily trounce humans when it comes to doing simple arithmetic. This has been true for decades if not over a century. But calculators obviously aren’t AGI. They can do one thing at a superhuman level, but are useless for practically anything else.
LLMs like chatGPT and Claude are more like calculators than AI hype-meisters would like to let on. When they burst onto the scene in late 2022, they certainly seemed impressively general. You could ask them a question on almost any topic, and they’d usually give a coherent answer so long as you excused the occasional hallucinations. They also performed quite well on human measurements of intelligence, such as college level exams, the SAT, and IQ tests.. If LLMs could do well on the definitive tests of human intelligence, then certainly AGI was only months or even weeks away, right? The problem is that LLMs are still missing quite a lot of things that would make them practically useful for most tasks. In the words of Microsoft’s CEO, they’re “generating basically no value”. There’s some controversy over whether the relative lack of current applications is a short-term problem that will be solved soon, or if it’s indicative of larger issues. Claude’s performance playing Pokemon Red points quite heavily toward the latter explanation.
First, the glass-half-full view: The ability for Claude to play Pokemon at all is highly impressive at baseline. If we were just looking for any computer algorithm to play games, then TAS speedruns have existed for a while, but that would be missing the point. While AI playing a children’s video game isn’t exactly Kasparov vs Deep Blue, the fact it’s built off of something as general as an LLM is remarkable. It has rudimentary vision to see the screen and respond to events that occur as they come into the field of view. It interacts with the game through a bespoke button-entering system built by the developer. It interprets a coordinate system to plan to move to different squares on the screen. It accomplishes basic tasks like battling and rudimentary navigation in ways that are vastly superior to random noise. It’s much better than monkeys randomly plugging away at typewriters. This diagram by the dev shows how it works
I have a few critiques that likely aren’t possible for a single developer, but would still be good to keep in mind when/if capabilities improve. The goal should be to play the game like a player would, so it shouldn’t be able to read directly from the RAM, and instead it should only rely on what it can see on the screen. It also shouldn’t need to have a bespoke button-entering system designed at all and should instead do this using something like ChatGPT’s Operator. There should be absolutely no game-specific hints given, and ideally its training data wouldn’t have Pokemon Red (or even anything Pokemon-related) included. That said, though, this current iteration is still a major step forward.
Oh God it’s so bad
Now the glass-half-empty view: It sucks. It’s decent enough at the battles which have very few degrees of freedom, but it’s enormously buffoonish at nearly everything else. There’s an absurdist comedy element to the uncanny valley AI that’s good enough to seem like it’s almost playing the game as a human would, but bad enough that it seems like it’s severely psychotic and nonsensical in ways similar to early LLMs writing goofy Harry Potter fanfiction. Some of the best moments include it erroneously thinking it was stuck and writing a letter to Anthropic employees demanding they reset the game, to developing an innovative new tactic for faster navigation called the “blackout strategy” where it tries to commit suicide as quickly as possible to reset to the most recently visited Pokemon center… and then repeating this in the same spot over and over again. This insanity also infects its moment-to-moment thinking, from hallucinating that any rock could be a Geodude in disguise (pictured at the top of this article), to thinking it could judge a Jigglypuff’s level solely by its girth.
All these attempts are streamed on Twitch, and they could make for hilarious viewing if it wasn’t so gosh darn slow. There’s a big lag in between its actions as the agent does each round of thinking. Something as simple as running from a random encounter, which would take a human no more than a few seconds, can last up to a full minute as Claude slowly thinks about pressing ‘A’ for the introductory text “A wild Zubat has appeared!”, then thinks again about moving its cursor to the right, then thinks again about moving its cursor down, and then thinks one last time about pressing ‘A’ again to run from the battle. Even in the best of times, everything is covered in molasses. The most likely reaction anyone would have to watching this would likely be boredom after the novelty wears off in a few minutes. As such, the best way to “watch” this insanity is on a second monitor, or to just hear the good parts second-hand from people who watched it themselves.
Is there an AI that can watch dozens of hours of boring footage and only pick out the funny parts?
By far the worst aspect, though, is Claude’s inability to navigate. It gets trapped in loops very easily, and is needlessly distracted by any objects it sees. The worst example of this so far has been its time in Mount Moon, which is a fairly (though not entirely) straightforward level that most kids probably beat in 15-30 minutes. Claude got trapped there for literal days, with its typical loop being going down a ladder, wandering around a bit, finding the ladder again, going back up the ladder, wandering around a bit, finding the ladder, going back down again, repeat. It’s like watching a sitcom of a man with a 7 second memory.
There’s supposed to be a second AI (Critique Claude) to help evaluate actions from time to time, but it’s mostly useless since LLMs are inherently yes-men, so when he's talking to the very deluded and hyperfixated main Claude he just goes with it. Even when he disagrees, main Claude acts like a belligerent drunk and simply ignores him.
In the latest iteration, the dev created a tool for storing long-term memories. I’m guessing the hope was that Claude would write down that certain ladders were dead-ends and thus should be ignored, which would have gone a long way towards fixing the navigation issues. However, it appears to have backfired: while Claude does indeed record some information about dead-ends, he has a tendency to delete those entries fairly quickly which renders them pointless. Worse, it seems to have made Claude remember that his “blackout strategy” “succeeded” in getting out of Mount Moon, prompting it to double, triple, and quadruple down on it. I’m sure there’s some dark metaphor in the development of long-term memory leading to Claude chaining suicides.
What does this mean for AGI predictions?
Watching this trainwreck has been one of the most lucid negative advertisements for LLMs I’ve seen. A lot of the perceptions about when AGI might arrive are based on the vibes people get by watching what AI can do. LLMs can seem genuinely godlike when they spin up a full stack web app in <15 seconds, but the vibes come crashing back down to Earth when people see Claude bumbling around in circles for days in a simplistic video game made for children.
The “strawberry” test had been a frequent concern for early LLMs that often claimed the word only contained 2 R’s. The problem has been mostly fixed by now, but there’s questions to be asked in how this was done. Was it resolved by LLMs genuinely becoming smarter, or did the people making LLMs cheat a bit by hardcoding special logic for these types of questions. If it’s the latter, then problems would tend to arise when the AI encounters the issue in a novel format, as Gary Marcus recently showed. But of course, the obvious followup question is “does this matter”? So what if LLMs can’t do the extremely specific task of counting letters if they can do almost everything else? It might be indicative of some greater issue… or it might not.
But it’s a lot harder to doubt that game playing is an irrelevant metric. Pokemon Red is a pretty generous test for many reasons: There’s no punishment for long delays between actions. It’s a children’s game, so it’s not very hard. The creator is using a mod for coloring to make it easier to see (this is why Jigglypuff’s eyes look a bit screwy in the picture above). Yet despite all this, Claude still sucks. If it can’t even play a basic game, how could anyone expect LLMs to do regular office work, for, say, $20,000 a month? The long-term memory and planning just isn’t there yet, and that’s not exactly a trivial problem to solve.
It’s possible that Claude will beat pokemon this year, probably through some combination of brute-force and overfitting knowledge to the game at hand. However, I find it fairly unlikely (<50% chance) that by the end of 2025 there will be an AI that exists that can 1) be able to play Pokemon at the level of a human child, i.e. beat the game, able to do basic navigation, not have tons of lag in between trivial actions, and 2) be genuinely general (putting the G in AGI) and not just overfit to Pokemon, with evidence coming from being able to achieve similar results in similar games like Fire Emblem, Dragon Quest, early Final Fantasy titles, or whatever else.
LLMs are pretty good right now at a narrow slice of tasks, but they’re missing a big chunk of the human brain that would allow them to accomplish most tasks. Perhaps this can be remedied through additional “scaffolding”, and I expect “scaffolding” of various types to be a big part of what gives AI more mainstream appeal over the next few years (think stuff like Deep Research). Perhaps scaffolding alone is insufficient and we need a much bigger breakthrough to make AI reasonably agentic. In any case, there will probably be a generic game-playing AI at some point in the next decade… just don’t expect it to be done by the end of the year. This is the type of thing that will take some time to play out.
TheAntipopulist
Formerly Ben___Garrison
5mo ago
Nice book review.
I too wonder why prostitution or sex tourism is still so shunned. It's clear why the far left and far right hate it: the Fascist-Feminist Synthesis holds that women have no agency in such a situation, and that they must be protected from their own decision to offer themselves to beastly men.
But why does the center go along with this still? Residual Puritanism might explain some part, but I doubt it's the whole answer.
TheAntipopulist
Formerly Ben___Garrison
2mo ago
Last week there was a conversation on here about a potential peace deal in Ukraine. I claimed that the peace deal seemed fake since if you knew the background on peace efforts, you'd know that both Putin and Zelenskyy were playing a goofy game trying to pin the other one as the one who "doesn't want peace" in the eyes of Trump to try to direct Trump's ire in the other direction.
We now have pretty good confirmation that no peace deal will be forthcoming in the near term. JD Vance has said that the war won't end anytime soon. This backs up further reporting following the mineral deal that Trump's team was looking for ways to compel Russia to come to the table, and didn't really find any options that they liked.
The bull case for a Trump-brokered peace deal was the idea that the US could use its power to demand that both sides come to the table, and if either side tried to walk away then the US could force them back. This worked halfway, as the US has a lot of leverage over Ukraine for things like intelligence gathering, air defense, and to some extent other military deliveries. Much of MAGA hates Zelenskyy personally, and Trump was more than willing to exercise that leverage when Zelenskyy snubbed him at the WH meeting. The problem was that the other half of the puzzle was missing. Some claimed that the US could threaten Russia by promising to "drown Ukraine in weapons" if Russia didn't come to terms. However, Trump has been unable or unwilling to do this, so we had the situation where Trump could compel one side quite effectively, but when the other side did something Trump didn't like all he could do was tweet "Vladimir, STOP".
Peace is good as a general rule, and it would have been good if Trump could have gotten a peace deal along the lines of "ceasefire at current lines of control, Ukrainian defense guaranteed by Europe" so it was worth a shot. But alas, it seems like the war will continue.
TheAntipopulist
Formerly Ben___Garrison
1yr ago
Not gonna happen.
Nobody seriously defends the superstitions of Christianity, and while social movements can survive a large amount of inconsistency, all attempts at "cultural Christianity" have basically been failures due to the inherent contradictions. It's just a bridge too far to try to harness the culture of Christianity while ignoring the superstitions that underpin them, while also having people who do believe the superstitions loudly proclaim they're undeniable truths.
TheAntipopulist
Formerly Ben___Garrison
7mo ago
People have at least discussed this, although I don't know how much it's been internalized yet. Matt Yglesias had an article about the crank realignment, Hanania had an article about voters who see conspiracies everywhere, and Meskhout had this article.
In short, both sides have become dominated by delusional partisans screaming in echo chambers. The left have become experts in infiltrating institutions and corrupting them to woke ends, while the right have become eternal dissidents who are great at critiquing the left but terrible at actually building better replacement institutions. The left was a bit ahead of the right when it came to radicalizing, but it's also deradicalizing now in a way that will likely happen to the right in a few years. Around 2020 was "peak woke" after which things slowly calmed down. Now we're approaching the summit of "peak crank" on the right, which will also hopefully calm down.
MAGA experienced a wave of euphoria from Trump's election until about around the time of the trade disputes. They felt like they were on top of the world, and that nothing could stop them. They notched a few wins against wokeness, but their major victory was in the realm of vibes.
It's increasingly seeming like those days are over. Scott Sumner's article details who's up and who's down over the past few weeks:
Who's Up:
-Neoliberals
-The experts
-TDSers
-The elite media
-Chinese and Canadian liberals
-Deficit hawks
-Principled conservative free speech advocates
-Integrity hawks
-Rules hawks
-Critics of bullying
Who's Down:
-Mercantilists
-The populists
-Anti-anti-trumpers
-Fox News
-Non-US nationalists
-Deficit doves
-Unprincipled conservative free speech advocates
-Issues people
-Autocracy advocates
-American exceptionalists
Edit for more opinions per moderator request: I agree with this article that the vibes have definitely shifted, as it's been clear in my (adversarial) conversations with MAGA that the mood has changed from combative (pre-election) to triumphalism (post election until a few weeks ago) and then back to combative with a hint of disillusionment (today). Any opposition movement is going to have principled believers and cynics, e.g. people who think we should have free speech as a general rule and people who only claim to like free speech but really want to censor their opponents when they come into power. Winning means these splits that could be swept under the rug get blown out into the open, and the pendulum starts swinging back the other direction. Hopefully we don't swing back to crazy wokeness, but I'd pretty much take any alternative at this point. A decade ago I would never have seen myself cheering for The Experts or The Media, but I've seen the alternative now, and it's just so much worse.
TheAntipopulist
Formerly Ben___Garrison
1yr ago
Sounds a lot like how a rationalist would approach a topic, no?
Yes, it indeed sounds so much like a rationalist that it also sounds like he's not defending the superstitions of Christianity at all!
I should have said "nobody relevant defends the superstitions of Christianity", i.e. there might be some, but no major public intellectual does it and gains any sort of traction. Ayaan Hirsi Ali was the most prominent defense in recent times which got a fair degree of traction, and not a single sentence in her defense was about the actual superstitions.
TheAntipopulist
Formerly Ben___Garrison
1yr ago
A better response to any misgivings about the FBI is to have R politicians probe the organization and gradually escalate if it's found to be breaking the law to assist Dems. On the other hand, saying "the enemy's misdeeds justify our own lies" is pure toxoplasma.
TheAntipopulist
Formerly Ben___Garrison
1yr ago
This was by no means a good debate for Trump, at least in terms of his individual performance. He blatantly refused to answer some questions, and they were the pretty important questions involving stuff like Putin and J6. Others were bad for different reasons, e.g. "will you accept the results of the election" with his response being basically "not if I lose". Worse than any of that though is that his responses were just incoherent. It's like if a lobotomized chatGPT was told to act like Trump, and it spewed a random collection of things that individually sounded like something Trump would say, but without any coherent structure or chain of logic. Trump has always had a meandering speaking style, but compare his performance tonight to the debates in 2016 and there's a world of difference. He's pretty clearly suffered a substantial age-related mental slowdown since then.
Of course, Biden's performance was way, way worse so it'll likely just be forgotten.
TheAntipopulist
Formerly Ben___Garrison
1yr ago
This story is a great encapsulation of two important phenomena:
- How utterly asleep at the wheel most Europeans were in regards to Russia, especially post-Crimea.
- How much more dangerous Russia could be if they got a handle on corruption. But alas, no dictatorship can really solve corruption since it's too beneficial to the leader at the top for maintaining his position.
Suppose communism is bad. How do you teach normies this?
You poisoned derailed the discussion by leading with this. Almost no normies actually think communism is good, nor are they yearning for it to any great degree. At worst they have some uninformed ideas that, if you squint, can sort of seem communist-adjacent. Stuff like supporting price ceilings or floors in competitive industries. But even these aren't really doing much damage. Things like "building more housing leads to higher housing + rent prices" has been much more disruptive to a flourishing society, and it doesn't spring from anything related to communism, but rather from ignorance of basic economics.
TheAntipopulist
Formerly Ben___Garrison
1yr ago
Huh? How snail-brained are 22-34% of these voters? Why would you care if he gets convicted?
The fact you'd say this is a pretty emblematic of how crazy the US (and this site) have become. This might seem like a hot take, but people generally don't want their leaders to be convicted felons.
Well, maybe that would have held more true back before trust in institutions collapsed. Those 22-34% are the last vestiges of that era. The thought is that anyone can lob an accusation, but a conviction carries more weight. Yes, most people understand that prosecutors would generally only bring cases that have a good chance of winning, but they can still fudge around the edges.
Nowadays, Trump could probably murder someone on live TV and a majority of the Republican voters would say he didn't do it. That's basically what the election loss denialism came down to. Why let evidence get in the way of vibes and dunking on the outgroup!
TheAntipopulist
Formerly Ben___Garrison
1yr ago
On a whim I decided to watch a bit of the Ben Shapiro interview, and I'm thoroughly unimpressed. When Ben asked him what his favorite argument is for the proof of God, he says what essentially boils down to the First Cause argument, something that's been trounced in the internet atheist debates for decades. When pressed with a follow up of what caused God then, he responded with special pleading. He dressed it up with fancy words like "that which is properly unconditioned on this reality", and his presentation is polished, but he's just regurgitating arguments from a debate that was largely settled over a decade ago. After watching a bit more and hearing nothing but a few "God of the Gaps" arguments I closed the tab.
TheAntipopulist
Formerly Ben___Garrison
2yr ago
This will likely accelerate the Left's attempts to move to a more pro-censorship platform. They also might pressure advertisers to boycott Twitter.
TheAntipopulist
Formerly Ben___Garrison
2mo ago
I covered that in my post that I linked. The notion that the bill was "open borders up to 5000 migrants per day" was just egregiously false.
After Biden's pardons and now this, it's pretty clear that the President's unilateral ability to grant clemency to anyone ought to be dramatically reduced, if not removed entirely. Literally every president in my lifetime has abused this power, and the expected guardrails (voters will punish bad pardons) mostly don't work.
There's nothing to stop a president from goading political violence or corruption, and then pardoning afterwards (or hell, even before!). It's a highly abuseable, very obvious point of failure.
TheAntipopulist
Formerly Ben___Garrison
3mo ago
Arguing from miracles is just... painfully bad. If you have strong evidence that could be tested and perhaps replicated of supernatural phenomena occurring on Earth, that would be one thing. But this is like debating Trumpian 2020 election skeptics, where they're full of reasons to sneer and hate their outgroup, but if you ask them to make a positive case for their own arguments, they wither and try to deflect. The best evidence I can think of to dismiss these people as a group is the fact they've failed to find a single good example to rally around (be it an example of election fraud that was widespread enough to make a difference, or a miracle that genuinely occurred). They all have their own little gish gallop of bad reasons that primarily rely on the audience not being familiar with the arguments, because any evenhanded analysis would show their points are bunk.
TheAntipopulist
Formerly Ben___Garrison
2mo ago
This wasn't true to any serious extent, other than how laws are always interpreted by the judicial system
TheAntipopulist
Formerly Ben___Garrison
1mo ago
I agree that this stuff is becoming more and more difficult to tell apart. We even had one of our own posters get falsely accused by the mods of using AI recently. People are going to claim many things are "obviously AI" when they actually aren't, and the mania of false accusations is going to tick a lot of people off. When you're accused of using AI, not only are people saying you're committing artistic fraud, they're also implying that even if you aren't then your output is still generic trash to some extent.
I wish the Luddites would go away and we could all just judge things by quality rather than trying to read tea leaves on whether AI had a hand in creating something.
This also 100% applies to this forum's rule effectively banning AI. It's a bad rule overall.
Capgains taxes are fine, and even desirable if you want to lower or stabilize the gini coefficient. Rich people tend to get most of their money through their existing wealth, not through directly working which would be subject to income taxes. It's the closest thing to a tax on wealth that most societies can really achieve. A society that lets wealth accumulate unhindered ends up looking like France in the Belle Epoque period, where dynasties of the ultra-wealthy control almost everything.
You can say something similar for a sales tax, where post-tax money is taxed again, and if inflation happens then the absolute value of the tax increases. None of this makes either tax "unfair" or "absurd".
I do agree that trying to fix the problem by subsidizing housing for the young is silly. It's treating the symptoms instead of the cause, which is almost certainly NIMBYs and zoning restrictions like it is in the USA. But those are typically local issues that the national government doesn't have jurisdiction over, so they try to seem like they're "doing something" by just throwing money at the problem.
It's really a scheme to tax old people and give the benefits to younger people, which isn't the worst idea but the underlying issues of the housing crisis really do need to be resolved as well.
More options
Context Copy link