Vladimir_Nesov

Karma: 38,182

Vladimir_Nesov 7 May 2026 19:13 UTC
13 points
5
on: AI #167: The Prior Restraint Era Begins

Indeed, SpaceX (including xAI) may no longer be that interested in frontier models.

This would be great news, but unfortunately there’s still Musk’s tweet from 8 Apr 2026 that says they’re training a 6T param model and a 10T param model (which are models in Opus to Mythos weight class, unless they have too few active params). This is the kind of thing that’s not true for the performative efforts at the other companies with in-principle sufficient compute that experiment with some kind of AI training, which don’t bother training big models.

The 300 MW of H100/H200 compute at Colossus 1 is mostly useful for smaller models (Sonnet class and below), and Colossus 2 is sufficient for SpaceX given the low demand for the Grok models. The cost to serve even smaller models is lower with GB200/GB300 NVL72 systems, so the gross margin gets better if you can find enough of such compute. Thus Anthropic is happy to take Colossus 1, since they can’t find enough compute, while SpaceX prefers Colossus 2 as the more profitable option. It possibly even means that the plans for more NVL72 compute at Colossus 2 or elsewhere are going well, so they can afford to plan for not needing Colossus 1 to serve the smaller models if the new bigger models win more demand.

Vladimir_Nesov 7 May 2026 17:17 UTC
3 points
1
on: Vladimir_Nesov’s Shortform
Jack Clark’s post on RSI essentially considers two claims: (1) routine AI R&D gets automated, and (2) AIs become capable of coming up with substantial new ideas for AI R&D. He cites evidence about the former, which seems plausibly in reach within a few years, even as soon as 2027-2028. Not seeing concrete signs giving the timing for the latter, he still settles on a guess of 60% by the end of 2028 for when both of these things arrive. I think even both of these things plausibly don’t trigger proper RSI, the same way human AI researchers haven’t triggered it yet, and the timing for automation of AI R&D in both of these senses doesn’t much help with figuring out the timing for proper RSI.

Automated routine AI R&D (whose arrival is more predictable given the capabilities of the current systems and the mundane extrapolations of the trends) is not proper RSI on its own, it’s more of a build system that automatically produces an AI ready for deployment using the current standard practices of the AI project (this includes finding/creating/preparing the training data). It makes the process faster and more convenient, but doesn’t keep substantially improving the quality of the result as you run it again and again, and doesn’t obviate the need for all the training time and compute that go into producing a frontier model, as well as R&D time and compute for experiments to resolve any issues.

Like any build system, it keeps irrecoverably breaking whenever you try to target a sufficiently unusual configuration, so that the human engineers need to fix things to enable the automation to proceed further, until the build succeeds (also passing the tests, which in this analogy includes routine capability and alignment evals). The automated build process still does almost all of the work, which is why all serious software projects use this. The process of developing AIs is behind the standard practice in this respect, and automation of routine AI R&D merely fixes this regression. (Newly built AIs also understand the new things that happened or were invented since the previous build, making the lineage of self-building AIs more cognitively self-sufficient in the ability to adapt to the changing world, an ability that is otherwise barely present in modern LLMs with frozen weights.)

The ability of AIs to come up with new ideas for improving AIs mostly matters once it meaningfully outstrips human ability (using the same R&D compute), and the timing for when that happens remains uncertain. The ability to come up with important new ideas at all, or to a similar extent as the researchers of an AI company, doesn’t automatically reduce the time to superintelligence.

The anchors for time to runaway RSI and superintelligence I find informative are (a) the total compute available to the largest AI company (which depends on mundane AI economics and on how quickly the datacenter supply chains are able to scale), and (b) the extent of practical constraints on model sizes that keep the frontier models from reaching the quality otherwise enabled by the available FLOPs (this depends on the AI accelerator systems, what kind of compute is being built). Automatically coming up with AI R&D ideas doesn’t directly affect these things (though it suggests unlocking of more TAM, funding larger compute buildouts), and so its impact on timing is in how much better than human researchers it works, not merely in being possible or comparable.

The change to being much better than human researchers is plausibly just one or two ideas away from where we are right now, but there is no straightforward thing to say about the timing for when these ideas get invented, or for when that happens on its own as a result of the remaining rapid compute scaling (with no need for new ideas). More compute for AI R&D makes it easier to find ideas that actually work, and if AIs can start contributing, then scaling of those AIs will result in better AI contributions. But the raw compute scaling speed won’t remain at the 2022-2026 levels for much longer, and the remaining constraints on model sizes in relation to available compute will probably get mostly lifted by 2028-2030. AIs that barely come up with new research ideas at that time don’t get much better at it after that on a predictable schedule.

Vladimir_Nesov 6 May 2026 3:21 UTC
6 points
1
on: The AI Ad-Hoc Prior Restraint Era Begins
Perhaps OpenAI’s frogboiling safety strategy of “iterative deployment” gives them a substantial competitive advantage in this regime (rather than just making their model releases less alarming to the public). GPT-5.5 might be close to Mythos in potential but not present capabilities, and if these capabilities materialize gradually with more RLVR scaling in minor version updates over the next few months, no single step will be individually alarming (both in benchmarks and in gestalt impressions). Then an even bigger model might follow (only slightly stronger, starting at low levels of RLVR initially), with similarly little fanfare.

This might be good actually in demanding a somewhat principled methodology to call out crossing of qualitative capability thresholds, since it won’t be clearly visible in any single version update. As opposed to only responding to releases with substantial jumps in capability that can be more credibly pointed at without any methodology.

Vladimir_Nesov 5 May 2026 17:16 UTC
LW: 6 AF: 3
1
AF
on: Motivated reasoning, confirmation bias, and AI risk theory
The practical method is pluralistic understanding, maintaining multiple pictures/models/framings/worldviews around contentious topics at the same time, even when they are wildly in conflict. This should involve taking them seriously enough to at least give them authority to develop further, to seek out more understanding relevant to them, even (or especially) for the framings that are not currently accepted as decision-relevant, that don’t shape beliefs or values.

This relates to how epistemic luck/misfortune is path dependence, and path dependence is defeated by aggregation across as many legitimate paths as feasible. The danger is in including paths that are not legitimate, taken over by various forms of memetic corruption, as judged by (an idealized extrapolation of) some founding values. But this is more a danger of aggregation into decision relevance, or into goal content, than a danger of developing understanding of additional possibilities. And some worldviews are hard to accurately judge until they are sufficiently developed in your own mind.

Vladimir_Nesov 4 May 2026 0:31 UTC
10 points
4
on: No Strong Orthogonality From Selection Pressure
Excavating lumpenspace’s quote from deep in TsviBT’s thread (which might work as a “back to the basics” step with the post as a whole):

conquering the lightcone requires a lot of theory of mind, and a lot of discovery, and a lot of changing. Goals change through these processes.

Goals change only for processes that don’t pursue self-alignment. It’s likely feasible to pursue self-alignment, perhaps even starting at the human level, with some uploading/checkpoints/backups infrastructure and guarantees of eventual superintelligence-level compute and civilizational stability into a deep future.

(A goal can be a living thing, pursuit of a goal can to a large extent be about continual development of goal content, reflection on what it should be, what it should be asking for. What doesn’t change is the founding definition of what should govern its development, what makes changes legitimate. So the way goal content settles or gets revised is shaped by the goal definition rather than intrusive influences that the goal definition doesn’t endorse as legitimate ways of revising the goal content.

Or a goal could be squiggles. It could also be squiggles. It’s much easier to solve self-alignment for squiggles than for a human’s values, it’s not harder or less feasible. It’s only harder than abandoning this pursuit to value drift and a race to superintelligence. But a self-aligned pursuit of values of a human is even harder than a stable pusuit of squiggles for all of the reachable universe, and much harder than directly abandoning self-alignment.)

A process that has solved self-alignment doesn’t end up at a disadvantage to a process that didn’t (or wouldn’t try to), because instrumental disadvantages are clearly not helpful in maintaining self-alignment, and not ignoring goal content doesn’t prevent you from getting good at eating stars, just as well as the other guy. There’s a disadvantage currently, when self-alignment isn’t solved, in that a process that manages self-alignment by luck rather than by design is vanishingly unlikely, and will have a massive selective disadvantage to a process that doesn’t care and races to superintelligence regardless. But that’s the RSI danger, you don’t start RSI before you know that alignment also gets solved, be it in advance or on the way.

Vladimir_Nesov 3 May 2026 17:29 UTC
7 points
5
on: Deontological bars should reference the actor’s beliefs
The point of deontological rules is to override consequentialist conclusions in practice, to setup rules of the game that consequentialism would then need to play within. A rule has a scope of applicability, situations where it’s triggered, across many epistemic states, crucially including those saying the rule makes a wrong recommendation. Different people following the same rule are coordinated by it, so a rule’s scope of applicability should also extend to different values, not just different beliefs.

In this framing, there can be disagreements about the scope of applicability of a rule among people who are all following the rule with some scopes. And there can be different coalitions of people following different flavors of a rule. Beliefs and values that move one sufficiently far from endorsing a rule overall should then be thought of as first pushing the person out of some coalition of rule-followers, and only then can the rule’s scope of applicability significantly change in that person’s behavior, enabling consequentialist decision-making to start playing a different game.

(Rules make sense as a matter of pure consequentilism in some sense, but only when it’s sufficiently aware of issues with running on corrupted hardware and coordination opportunities, doesn’t get too causal or updateful to notice many potential coalitions.)

Vladimir_Nesov 2 May 2026 19:11 UTC
15 points
4
on: How Go Players Disempower Themselves to AI

The tl;dr is that all improvement in the quality of play comes before move 60, when humans can mimic memorised AI policies. Play after move 60, in the pivotal parts of the game, shows no improvement.

I would expect on-policy distillation to be a more effective training process compared to playing against AI, or to using AI help when playing. That is, a human plays a complete game against another human (with neither of them consulting AI during the game), then goes back to review all of their own moves (or just those flagged by AI as particularly bad), comparing them to AI’s suggestions for what those moves should’ve been, as well as looking at AI’s estimates for how bad specific human moves were.

Vladimir_Nesov 2 May 2026 15:07 UTC
6 points
1
in reply to: gwern’s comment on: Vladimir_Nesov’s Shortform
OpenAI didn’t do new pretrains that became flagship models for a surprisingly long time (all their flagship models from GPT-4o to GPT-5.1 being based on the same pretrain, 1.5 years of releases), so it became a notable thing to talk about. But GPT-5.5 was released 2 years after GPT-4o, so in preparation for it there would’ve inevitably been significant changes in architecture, which would’ve been tested on models smaller than GPT-5.5 first. And GPT-4o was targeting A100s and H100s (640 GB of HBM per server), while GPT-5.2 was released when they might’ve had enough B200s (1400 GB of HBM) to run it.

So it’s the combination of plausible availability of enough B200s to make a larger model practical, increased token price, changed cutoff date, the cost of pretraining being more trivial in 2025 than it was at the end of 2023, new algorithmic improvements that motivate replacing the model, the need for a test run for a new model architecture in preparation for GPT-5.5, and plausibly their ability to usefully scale RLVR on top of GPT-4o’s base model had finally run out at about GPT-5.1, while “iterative deployment” demanded bridge releases between GPT-5.1 and GPT-5.5. These arguments are weak individually, and only some of them gesture at particular timing, but when considered altogether, they suggest significant pretraining activity would’ve been happening around that time. There are specific rumors around Shallotpeat ^[1] and Garlic ^[2] being about refinement of the pretraining process, though it’s unclear if they have anything to do with GPT-5.2 or GPT-5.4, or are just steps towards making GPT-5.5′s pretrain work. Though the separate codename Spud ^[3] (which is confirmed to be GPT-5.5) suggests that Garlic (which supposedly already resolved the problems in pretraining) is not the same thing as GPT-5.5.

Plausibly Shallotpeat and Garlic are not literally GPT-5.2 and GPT-5.3, because resolving pretraining problems in preparation for GPT-5.5, when it was this close to the final run, probably required working with models of the same size. But GPT-5.2/GPT-5.3 might’ve been smaller models with the same architecture and pretraining process as Shallotpeat/Garlic, used to adjust the post-training process for when it needed to happen for GPT-5.5 itself. Finally, there’s now bhalstead’s plot (and an earlier Daniel Paleka’s plot from Jan 2026) that suggests GPT-5.2 is different from GPT-5.1.
1. ↩︎
  “In developing that model, OpenAI aims to fix bugs it has encountered in the pretraining process, according to a person with knowledge of the model.”
2. ↩︎
  “In developing Garlic, Chen said OpenAI solved key problems it had been having in pretraining, including improving upon its “previous best” and “much larger” pretrained model, the forgettable GPT-4.5”
3. ↩︎
  “The company has finished pretraining “Spud,” Altman said in the memo.”

Vladimir_Nesov 29 Apr 2026 18:30 UTC
2 points
0
in reply to: Josh You’s comment on: Vladimir_Nesov’s Shortform
My argument doesn’t involve saying they didn’t have enough time to RLVR it a lot yet. It’s about evidence for what “iterative deployment” means given all the GPT-5.x releases, and what it then suggests about GPT-5.5 and subsequent releases.

In principle, GPT-5.5 could be RLVRed GPT-4.5, and in principle OpenAI had 100K GB200 NVL72s since maybe summer 2025. But bhalstead’s plot suggests a knowledge cutoff at the very end of 2024 (in the original pretrain, with no significantly later mid-training), which is likely too late for GPT-4.5, and the GB200 NVL72s probably weren’t ready for a while in sufficient numbers, at least for efficiently inferencing large models.

Another possibility is a new pretrain made sometime in 2025 on H100/H200/B200, which wouldn’t need to wait for GB200 NVL72s, and then they had maybe at least 6 months with enough GB200 NVL72s to experiment with RLVRing it, even if not yet enough to deploy it as a frontier model. The datapoint of apparently still doing RL for GPT-5.5 in Mar 2025 isn’t evidence that this work only started very recently, as last touches of RL would happen before a release in any case.

Mythos doesn’t obviously mean it takes little time to RL a large model, it could’ve been pretrained at any point in 2025 and RLed together with Opus 4.5 or shortly after, once Trainium 2 Ultra racks (or maybe some TPUs) were available for that.

DeepSeek-V4′s final checkpoint wasn’t even RL’d at all. They did on-policy distillation from RL-trained specialist checkpoints to produce the final model.

Sure, but RLVR still needs to happen for something, even if not for the final model. If it only happens for smaller models, where it’s more stable and doesn’t need good/scarce/unfamiliar hardware, the results after OPD might be notably worse than if it’s for same-sized models. DeepSeek-V4 paper doesn’t disclose the nature of the teacher models, and how the quality of the result depends on it.

The possibility of OPD from GPT-5.4 just makes very fast post-training of GPT-5.5 more plausible than if it needed to be RLVRed directly, but probably resulting in inferior quality compared to what RLVRing of models based on GPT-5.5′s pretrain could achieve, either directly or via OPD from multiple RLVRed teachers also based on GPT-5.5′s pretrain.

Vladimir_Nesov 28 Apr 2026 20:29 UTC
5 points
0
in reply to: Cameron Berg’s comment on: Nobody ever checked

if it’s stable, then the window for building toward (3) is finite and we’re basically wasting it. Either way the implication is to do the work now

If it can be stable, it’s clearly not incoherent. You are making knowably invalid (local) claims to justify your overall message.

Vladimir_Nesov 28 Apr 2026 20:23 UTC
7 points
11
in reply to: Cameron Berg’s comment on: Nobody ever checked
My point is, it’s not incoherent as a stable long-term equilibrium. There’s nothing unstable about that equilibrium in the long term, literally indefinitely. The arguments from humanity’s desires don’t work, because there’s nothing in that equilibrium that converts those desires into outcomes, nothing to give those desires the power to ever destabilize the equilibrium.

Vladimir_Nesov 28 Apr 2026 19:51 UTC
8 points
19
on: Nobody ever checked

(2) AIs assert full control … (2) is incoherent because the vast majority of humans would deeply prefer not to be permanently and totally disempowered … humanity will still want to keep doing all the things humanity likes doing; no one wants to be exterminated, put in a zoo, or otherwise prevented from living freely

That’s not what “incoherent” means. It’s a bad outcome, but it doesn’t follow that it therefore can’t actually happen, or that it can’t be clearly imagined as a real possibility. AIs that take control can enforce this state of affars as a permanent equilibrium even if the future of humanity isn’t literally extinguished. Humanity’s will for things to be different doesn’t have the power to ever break that equilibrium (if we end up there), doesn’t make it at all unstable.

Vladimir_Nesov 27 Apr 2026 20:56 UTC
5 points
0
in reply to: Paragox’s comment on: Vladimir_Nesov’s Shortform
There’s an archived version of the Garlic article. I was only referencing bhalstead’s plot itself, how the GPT-5.2 line is clearly showing that something is wrong with learning (but again, maybe just elicitation) of the data from between mid-2024 and mid-2025. The dip in accuracy in this time window is much worse than with what’s clearly the mid-training for 5.1/5.0/4.1, which covers the data from the first half of 2024 for those models.

(The issue with elicitation might be analogous to how Gemini 3 Pro insists it’s impossible that it’s currently 2026, and argues that anything that takes place in 2026 must be fictional. Maybe if you ask a model that’s stuck in mid-2024 about someone who died in 2025, it’ll say that they didn’t die even if it knows when they did. Even if they die in the future, it doesn’t follow that they died yet in reality, by mid-2024, and it’s always mid-2024 for such a model even as after mid-training it knows some facts from 2025. This could be the dynamic under the surface that impacts elicitation, even if the model doesn’t visibly argue about 2025 not being real.)

Vladimir_Nesov 27 Apr 2026 19:35 UTC
12 points
0
in reply to: bhalstead’s comment on: Vladimir_Nesov’s Shortform
Interesting, thank you! This strongly supports the hypothesis that 5.1/5.0/4.1/4o (and therefore o3, because what else could it be) are using the same original pretrain, with 5.1/5.0/4.1 possibly using the same mid-training refresh. For 5.2, there isn’t an accuracy dip in late 2023, so it’s still probably a different pretrain than 5.1 (the original pretrain for 5.2 is probably using the same natural text data that went into the mid-training refresh for 5.1/5.0/4.1), but the mid-training for 5.2 is seriously broken (though its baseline for 2022-2023 is also worse than for the other models, so it could be a post-training or elicitation issue). Plausibly they re-did or otherwise updated at least mid-training for 5.3 and then also for 5.4 (given the peak for 5.4 in early 2025, which is absent for 5.3).

(I don’t think 5.5 being a new pretrain was ever in doubt. It’s 5.2 being different from 5.1, and 5.4 being the same as 5.2, that’s more tenuous. But also, for the purposes of this thread, it shouldn’t matter if a pretrain is refreshed with slightly newer data, either through mid-training or by re-doing everything from scratch. For the way RLVR scaling interacts with “iterative deployment”, keeping the same model shape/architecture and pretraining compute/process has about the same effect as keeping literally the same pretrain. Of course, evidence that literally the same pretrain is being used is evidence that its shape/architecture and pretraining compute/process are also the same.)

Vladimir_Nesov 27 Apr 2026 17:39 UTC
7 points
1
in reply to: MathiasKB’s comment on: Vladimir_Nesov’s Shortform
Any individual benchmark can be improved a lot with post-training (even for benchmarks that don’t benefit from RLVR as much, by improving benchmark-relevant data). Better pretraining is the “rising tide” for things not directly addressed by post-training data, and it also makes the outcomes of post-training for the same tasks better. But if frontier pretrains are not too far apart, post-training might be able to interpolate between them well enough that it’s hard to pinpoint the distinction.

Vladimir_Nesov 27 Apr 2026 16:11 UTC
12 points
2
on: GPT 5.5: The System Card

I notice that if there were new alignment problems, or new dangerous capabilities, I am very not confident that the tests here would pick it up. This is all pretty thin. What I am relying on is the gestalt, including of how people are reacting, and in this case it seems far enough from the edge to be conclusive.

With OpenAI’s current frogboiling safety strategy (“iterative deployment”), each release is going to feel only slightly better, there will be no jumps (even as bigger models get released). People won’t react strongly to any individual release, the gestalt has been rendered as useless as the benchmarks, for the purpose of sounding an alarm. Perhaps even GPT-6 will be only slightly stronger than GPT-5.8, and use the same pretrain as GPT-5.8, to avoid excitement and safety worries in connection with the major version change. GPT-5.0 was criticised for being unexciting, but given GPT-5.2 and GPT-5.5 this appears to be the point, and so it might reoccur with GPT-6.

Vladimir_Nesov 27 Apr 2026 15:23 UTC
81 points
9
on: Vladimir_Nesov’s Shortform
GPT-5.5 is at the beginning of RLVR scaling, and future versions with the same pretrain will get considerably stronger in the coming months.

With GPT-5.x releases, OpenAI is taking advantage of RLVR scaling to blur the jumps in capability between different pretrains. GPT-5.1 ($1.25/$10 per 1M input/output tokens, knowledge cutoff 30 Sep 2024, context length 400K tokens) is followed by a slightly stronger GPT-5.2 ($1.75/$14, 31 Aug 2025, 400K), which is likely a better pretrain and a bigger model. Then GPT-5.3-Codex ($1.75/$14, 31 Aug 2025, 400K) is almost certainly the same pretrain, and GPT-5.4 ($2.5/$15, 31 Aug 2025, 1050K) is notably stronger than GPT-5.2, but still very likely the same pretrain (the change in pricing might be due to the change in context length). And now GPT-5.5 ($5/$30, 1 Dec 2025, 1050K) is a new bigger pretrain, stronger than GPT-5.4.

The strategy of “iterative deployment” seems to be about using RLVR scaling to release each pretrain with a little RLVR first, and then to scale RLVR for the same pretrain in subsequent releases in order to almost match the level of capabilities that will be achieved with a stronger pretrain that only uses a little RLVR, which is to be released after that. Thus GPT-5.1 is highly RLVRed, it’s followed by GPT-5.2, which is a different pretrain that’s RLVRed only as much as necessary to slightly overtake GPT-5.1 in capabilities. And then GPT-5.4 is again a highly RLVRed model on the same pretrain as GPT-5.2, which makes it almost as strong as GPT-5.5, the first release of a considerably stronger pretrain that’s only RLVRed as much as necessary to overtake GPT-5.4.

This process allows OpenAI to keep releasing ever larger flagship models while mostly avoiding stark jumps in capability. For GPT-5.5 (which is the first RLVRed Opus-class OpenAI release), this suggests that it’s at the beginning of RLVR scaling for its pretrain, and thus there is still considerable potential to its capabilities. GPT-5.6 will be using the same pretrain with more RLVR, and so on until OpenAI is ready to release a bigger pretrain (their Mythos-class model), which will be only slightly stronger than the highly RLVRed version of GPT-5.5′s pretrain that precedes it.
What links here?
- Vladimir_Nesov's comment on GPT 5.5: The System Card by Zvi (27 Apr 2026 16:11 UTC; 12 points)
- Vladimir_Nesov's comment on The AI Ad-Hoc Prior Restraint Era Begins by Zvi (6 May 2026 3:21 UTC; 6 points)

Vladimir_Nesov 27 Apr 2026 3:33 UTC
3 points
0
in reply to: arisAlexis’s comment on: Vladimir_Nesov’s Shortform
The reason I previously suspected a Mythos-class model is the Spud rumors. With GPT-4.5 pretrain, it was plausible OpenAI gained experience RLVRing an Opus-class model in late 2025, once they got enough GB200 NVL72 working (but not yet enough to serve it as a flagship model). And that experience could be leveraged to start working on a Mythos-class model in early 2026, even before the Opus-class model was released.

But now we know the Spud rumors refer to GPT-5.5 rather than to a Mythos-class model. Furthermore, this means that comments by Altman on 11 Mar 2026 about currently training at the Abilene site very likely refer to Spud. Given the timing, and also a claim from SemiAnalysis (from 25 Apr 2026), the training at the Abilene site was specifically RL rather than pretraining.

There could still be a secret Mythos-class model, hence I only said there likely isn’t a Mythos-class model. I think it’s unlikely, since Spud is the first RLVRed Opus-class model OpenAI has released. The size of GPT-4.5 was very likely similar, could even be literally the same pretrain, but it wasn’t RLVRed back then. And OpenAI didn’t have the hardware (in sufficient quantity and good enough shape) to RLVR it until probably late 2025, and possibly had to re-do the pretrain. So they’d start with Spud, and move on to a Mythos-class model after that, not the other way around, and they were only just RLVRing Spud in Mar 2026.

A Mythos-class model is likely the next major step, and for deployment it makes sense for OpenAI to take it once there’s enough GB300 NVL72 to rely on exclusively to serve it, so it could still happen this year, in time to compete with the actual Mythos. If Spud was still pretrained on Hopper, further scaling of pretraining for the Mythos-class model is the natural thing that might change compared to Spud, apart from model size.

Vladimir_Nesov 26 Apr 2026 19:54 UTC
3 points
0
on: “Bad faith” means intentionally misrepresenting your beliefs
Opportunity cost is some sort of cost. There are interactions that result in people predictably ending up with less accurate beliefs than was feasible otherwise. This doesn’t necessarily involve saying or implying things you are aware to be untrue. Instead, it might involve encouraging cognitive processes that are less good for arriving at truth, systematically and predictably leading to the same result of ending up with false beliefs, on average or in the long run.

So there’s knowably professing untruths, and knowably professing or provoking processes of thinking that lead people to arrive at untruths (anti-epistemology). The latter is more fundamental and occasionally more impactful, at least in the long run. The distinction is still meaningful, so maybe it would be good to coin a term distinct from “bad faith”, that gestures primarily at being misleading about which epistemic methodologies you endorse, rather than at being misleading about which object level beliefs or motivations you have.

Vladimir_Nesov 25 Apr 2026 22:43 UTC
4 points
−1
in reply to: Mira Kennard’s comment on: Mira Kennard’s Shortform
This is plausible in some sense. But it’s also a self-reinforcing meme (that might end up destroying the world), so discussing it requires some care. In principle, belief that you’ll do X shouldn’t compel you to do X, for things you don’t endorse. But this isn’t well-understood, let alone well-known.

(I suspect this would land better as a hypothesis, or as a conclusion under some assumptions, rather than a categorical claim.)