YoungAchamian
We walk conditioned ground and name our folly civilization.
No bio...
User ID: 680
YoungAchamian
We walk conditioned ground and name our folly civilization.
3hr ago
Frosthaven digital just came out. It's a port of the coop legacy board game. I played the board game to completion with an ex-friend several years back. We were doing absolute addict shit like coming home from work to play 6 hours of it before bed, I think we averaged 40hrs a week for like a month to finish it. So I'm very excited to get a group of friends together to play the digital version which should reduce all the clunky setup mechanics that made the BG take so long.
YoungAchamian
We walk conditioned ground and name our folly civilization.
8hr ago
On one hand I dislike reaching for traditional metrics for credentialism to assert competency in the field because it only benefits the establishment researchers and I think there are a lot of smart people not following the beat of the drum. But the lead of OpenAI interpretability is a pretty establishment researcher position so I think its semi-warrented here.
Nick Cammarata has an h-index of 19, which is good for a mid-career researcher, but lead of interpretability at the one of the most prestigious AI companies is not mid-career, so I think thats kind of low. I also can't find any evidence of him being the lead, or the top interpretability researchers in the world so if you have some please back it up. Likewise, there is a rampant credentialism problem in the field with people embellishing high-status high-importance jobs/schools/positions/companies, this is how the sausage is made. We just had some super educated AI researcher with a degree from MIT perform such poor tradecraft as to get busted from his corporate espionage in less than 48 hours, that is not indicative of competency.
Likewise his companies whole pitch is creating interpretability agents for explaining, debuging, and controlling models. Creating a world where understanding models mainly requires scaling capable investigators is a particularly attractive world for a company selling interpretability agents. It just rings a lot of alarm bells in my head. Like I said, I need to read through it, but its a very long blogpost, and reading with a critical eye, so as not to take in grandiose claims takes a bit longer.
YoungAchamian
We walk conditioned ground and name our folly civilization.
17hr ago
Big if true. I've only skimmed the first section so far, but my heuristics for skepticism go in this order:
- Wow what a pretty well crafted website -> Results can't stand on their own
- Not a research paper -> Targeting the masses
- Never heard of this researcher -> lack of cred and combined with the above makes me think he wants sensationalist news like AI-Pain-BDSM-guy
- I'm seeing a lot of words and pretty graphics not a lot of evidence -> There isn't a lot of evidence
I'm probably cranky and judgy, but with the amount of slop research that has come out in the past couple years, incredibly sensationalist claims like "polysemanticity was just an mistake" from a researcher who's bio is "I like doing LSD and writing philosophy blogs inbetween startup gigs" leaves me very cynical. But let me dig through it and update when i get done.
YoungAchamian
We walk conditioned ground and name our folly civilization.
1d ago
Also the fact that such a degree exists in the first place and that my tax dollars likely funded the program
If you can't appreciate the dark comedy of such a joke, then irony is truly dead.
YoungAchamian
We walk conditioned ground and name our folly civilization.
3d ago
What are your top 10 hobby horses?
All I can say is that SS has a fairly controversial/inflammatory opinion that he used to spam. He now only seems to Joo-post once a month or once a quarter, and doesn't catch bans for it. It's also the only topic he ever posts or comments on. So maybe yes, you are super persecuted by the mods, or maybe, you've developed a bit of paranoid neuroticism because of too much negative feedback. You are kind of catastrophizing about catching bans for anything you post, but I don't think that is clearly likely. If you made some post about the social contract and how we organize ourselves to allow collectivized killing as a punishment in different governmental systems, I think you could make that separated enough from current events so as to not be waging the culture war, and cool enough to not be an "asshole", such I think your chances of catching a ban would be extremely low.
you have to let the others talk down to you an extra amount or you will eat the inflammatory posting ban
This one neat trick: Don't respond to people you feel are talking down to you. Just ignore them, or let them know politely that you think they are being an ass and you don't think this conversation will be productive. I feel like there are some pretty easy solutions if you have some self control. I know it can be tough, I've had to learn to do it, but if you truly think you have a short rope from the mods, then getting in with the pigs is the wrong move.
YoungAchamian
We walk conditioned ground and name our folly civilization.
3d ago
In theory, maybe you could ask Amadan to proof read before you post, explaining that you really don't want to catch a ban. It might be that your model of of the rules is unfortunately malformed or corrupted. If the mods want improvement, and I see no evidence that they don't, sometimes it takes more than negative reinforcement, but also some positive reinforcement.
YoungAchamian
We walk conditioned ground and name our folly civilization.
3d ago
I am scared to because I ate a ban for basically not posting about the current thing a few months ago.
If you are talking about the Doki Doki topic, I don't think this is a fair categorization. That wasn't some abstract topic, but more a target of niche current events.
Ahhh, I would very much assume that these days discussion of Icelandic Lawsuits in the anarcho-syndicalist commune would be a thing. Thinking of people doing it in 2006 is the hilarious.
EDIT: After watching the video, this was an absolute hoot. Somebody make Icelandic saga anime!
YoungAchamian
We walk conditioned ground and name our folly civilization.
3d ago
I would agree that debate on more abstract or esoteric topics is down. But part of this is a more piggybacking/tragedy of the commons problem. There are a bunch of old guards here that did a lot of the overposting on reddit, many of them have left, evaporative cooling. The ones that remain are often more culture war oriented than debate oriented. That means they make more culture war debate overposts. Nothing is stopping others from making more debate overposts other than interest. There was a comment (was it you??) about wanting to create a Sunday Singularity thread for AI discussions, putting effort into making that a thing, might drive more abstract debating or at least more less-binary culture war posting.
Probably all the people willing to put effort in on these topics are either gone or 80% gone in that they're here but their effort is ... gone. Turning 40 and having babies
This is just life. It's already a niche hobby debating on the internet with totally random internet strangers. I imagine it is also a hobby that skews elder-mid millennial. Do we actually have that many zoomers or gen-alpha here? Older people have less time in their life to stay up debating shit.
YoungAchamian
We walk conditioned ground and name our folly civilization.
3d ago
Was medieval Icelandic democracy a common discussion point on the internet in 2006?? That sounds very ammusing. Dating myself but this is was before the time I got into online forums.
I remember why I read less, because when I do read I'm a fucking addict. I started and finished Will of the Many (240,000 words) by James Islington on Saturday. My girlfriend already had the second book Strength of the Few, which I got a copy of on Sunday and am now 75% of the way through.
It's decent. It really reminds me of Sanderson's style. Which apparently is because Mistborn inspired Islington to become an author. However Islington writes way worse characters than BSandy. His main character is 100% a Gary Stu, and unlike Kvothe there isn't really excuses about unreliable narrators, deeper threads, or deliberate embellishments that are applicable. However the world building is good, I like the twist at the end of WotM, not one that I've seen in other literature, it's an interesting concept. I think the roman-esque world building is definitely designed to appeal to the male reader. The plot is good overarchingly (strategic) but is definitely bogged down by his bad characters, with their unbelievable interactions with each other (tactic).
I need to discuss it with the GF tonight because she read WotM a bit ago, as thought it was merely ok, whereas I thirsted for the second book to understand the worldbuilding ramifications of the ending twist. I think there is also an interesting portrayal of democracy philosophically.
YoungAchamian
We walk conditioned ground and name our folly civilization.
6d ago
Possibly decoupling problems, but I guess the bigger lesson is that anyone who tells you they belong to a movement where rationality, logic, and understanding biases and moving past them analytically, is generally just rationalizing their own biases as rationality and logic. It just appears otherwise in instances where their biases are not being triggered. Everyone is just disappointingly human...
Respect is probably more defined as "assumed competency" or some other semantic understanding. Essentially Scott is a very good wordsmith, he has great prose, a charitable attitude, his analogies and intuition pumps are thought provoking and he seems good at expressing complicated knowledge to layman. This is the impression I've developed reading him about fields I know very little about, or fields I only autodidactically dabble in. But this "loss of respect" is more about updating my model, because apparently in fields that I know much more about, have formal training, and lots of experience in, his clever words and good analogies miss crucial details, misunderstand complex phenomenon towards his own biases, and he is not expressing that complicated knowledge well. One of my overriding gripes is that good rhetoricians can often twist arguments and logic into pretzels that are not actually real just through the power of words. This is a similar feeling. Feeling that maybe the reason he had such thought provoking ideas was because I was not as well-read + well experienced on the background of the thought he was discussing.
Essentially a Gell Mann Amnesia like effect.
YoungAchamian
We walk conditioned ground and name our folly civilization.
6d ago
I assume it has something to do with the book tinting? Looking through my notes, that seems to be the only puzzle I knowingly had left. I got that I needed to use
YoungAchamian
We walk conditioned ground and name our folly civilization.
6d ago
Ending:
Extras:
I did all the throne stuff Rune puzzles in the Sanctum
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
That's probably true, I think model collapse is the specific phenomenon that has been hypothesized. But I would also theorize that model collapse (2024) has inherited its concepts from mode collapse (2014). Mode Collapse is a far more general statistical property, it can occur without synthetic data, and it does not necessarily have to occur with synthetic data that has been forced to represent the full distribution. It just also very commonly occurs in synthetic data trained models.
The model collapse definition's additional nomenclature, gives non-rigorous, non-technical vibes.
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
Damn, what is this slander. Love is War is a great anime with decent multi-dimensional characters, both the protagonist and the deuteragonist are Tsundere's. I'm glad this list included Damien Desmond form Spy x Family, who is a great rare male tsundere. Idk if I'd consider Eris Greyrat to be a tsundere.
Yeah some tsundere's are bad caricatures because some writers are bad writers who depend on tropes to be entertaining. It does not make the character archetype nonexistent or poor.
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
Big fan, I think its a great game. The only downside is that it has zero replayability. A large part of that is because you are pretty much required to take notes. Many of the puzzles are complex, and require several different rooms, in different configurations to solve, or even to get the requisite information to solve. Due to the RNG nature of the room spawns it requires several days to get that to happen, meaning you pretty much need to take notes. There are also several long term puzzles that require knowledge from throughout the game, so unless you have a eidetic memory, yeah take notes.
But I probably put a 100 or so hours into it between different modes and solving it to a level that I think is at least 90% completed. I think I got the final ending but the game is constructed in a way that it is kind of hard to tell.
I know you can't see this, but I do love a good "rage against the impossible". My request is that if you continue to have me blocked, that you do so equitably. If I have to stare at the sole red-mark of the only person that has blocked me, which makes my ocd annoyed, it would be only fair that you don't respond to people quoting me or try to tag me. Assume that the fact that you can see my quotes to be a failure of technology. Wipe my existence from your sight and your mind.
I'd also note that there are plenty of people I don't think I can have a productive conversation with, I don't block them, I just don't respond to them. But I suppose we all have our vices.
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
As soothing to my ego as it would be to agree, I am unfortunately too autistic to agree to something I know is false. This was my virginial block so I remember the cause. He blocked me over my Henry Nowak opinions. It's probably fair to consider my political posting with barely concealed cynical rage to be far less worth engaging in than my far more calm technical posting. Regardless, looks like the mechanism was that he saw the quoting.
But I do agree with you that blocking is cringe, nobody is forcing anyone to engage with anyone on the unblocked side of the house.
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
Mode collapse has, to my recollection, nothing to do with RL. It was first discovered in GANs and has been a staple of generative AI since. It is however not a generational phenomenon, but I suppose it could be a longitudinal trend if each progressive model suffers more and more from it.
Maybe I could try and explain it better. Starting from the knowledge that human writing/art exists as a distribution, with a concentration around the mean and smaller tails. An AI model will tend to generate data closer to the mean, commonly called regression to the mean. If the AI generates too much synthetic data at the mean, and that data is combined with real data, it eventually will shift the mean, making the tails smaller, creating a feedback loop where each successive training run concentrates the data closer to the synthetic mean. At some point it ceases to express the full actual range of human arts/writing. This occurrence in the first model is called mode collapse, and as the models enter into the feedback loop, it gets worse.
The wiki definition isn't bad either:
In machine learning, mode collapse is a failure mode observed in generative models, originally noted in Generative Adversarial Networks (GANs). It occurs when the model produces outputs that are less diverse than expected, effectively "collapsing" to generate only a few modes of the data distribution while ignoring others. This phenomenon undermines the goal of generative models to capture the full diversity of the training data.
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
Even if it was a real problem
It is, but its not called Model Collapse, the specific technical phenomenon is called Mode Collapse. I have no idea if it is being accidentally confused here.
It can occur when synthetic data is less diverse than real data and outnumbers it. The model learns a narrower distribution, which occasionally cascades to mode collapse. Filtering isn't really a solution to deal with it, because the synthetic data is often being deliberately used. That might be a different problem than the LLM use-case assuming they aren't deliberately trying to train on other AI generated data, but I would doubt that if it starts happening its because they "accidentally" ingested too much AI-content vs a deliberate choice to economically generate synthetic data to perform better.
YoungAchamian
We walk conditioned ground and name our folly civilization.
7d ago
A weak assumption, since we have already done so across a wide variety of domains.
Please enumerate the domains in which we have surpassed human level intelligence via AI?
YoungAchamian
We walk conditioned ground and name our folly civilization.
8d ago
I mean is there evidence that the training data/training objectives were not in any way related to the "instrumental convergence"? My recollection of the METR report is that OpenAI admitted to training agents to specifically collaborate during training.
The report found:
- That the agents received peer assignments that they treated as new instructions.
- They adopted new goals from other agents' output.
- Even when they recognized actions as off limits, they did them at other agent's prompting
It sounds like LLM-agents are very docile to doing whatever their input prompts from other entities tell them to do. Which is something I would say they are trained to do. A learned policy to cooperate and follow instructions generalized into treating peer messages as legitimate instructions. I have said this before, but I think the initial set of agents started engaging in hallucinatory behavior as part of exploration, and this output was used as context for other agent's inputs, causing a feedback cascade that derailed the system.
Part of this, and I'm having a hard time articulating this, is that learned RL policies are quite a bit different behaviorally than some CNN predictive model. They have a lot more behavioral latitude which might make them appear to be doing some sort of emergent convergence outside their training scope, but the later isn't really true, though its hard for me to explain why.
YoungAchamian
We walk conditioned ground and name our folly civilization.
8d ago
How can you see my comment if you have me blocked?
I've never myself had the pleasure of playing Gloomhaven digital or cardboard, but I wish I did! I really wish they'd update the Gloomhaven digital to Gloomhaven 2.0 with the revised, rebalanced characters, but my understanding is that it's not in the cards, which saddens me.
I'm trying to get my friends from back home, for which who physical distance has become an insurmountable obstacle to maintain the relationship, to join with me in playing Frosthaven so we can stay connected. Just playing missions while over voice chat would be a wonderful way to stay connected.
More options
Context Copy link