I think the answer is straightforward. You complain that a substantial PR is no longer evidence of good faith. What is evidence? Participation in the community.
I think mailing lists (or an equivalent) are coming back. In a big way.
That xkcd is exactly what this thread is about, and so your lazy linkdrop is completely beside the point. The bots are currently posting lots and lots of (often superficially) "constructive and helpful" comments, and it turns to be a complete slog and burnout for humans to figure out whether to press the "constructive" or "not constructive" button.
I do dislike the fact that they nowadays use Discord though, there are better alternatives like fluxer.gg available, though I do understand that most/some people are already on discord and it has become more normal to create a discord server than other alternatives if one wants to which requires having members
which is much easier on discord.
As compared to the friction of another app even though they might be really great, I highly recommend trying out fluxer)
Then also if the developer has already created a discord server, then its hard to slightly migrate the existing users on their discord.
But I am sad about this development though, the community of open source or communities in general moving to discord does feel like moving to a walled garden (one which has increasingly more issues like trying to auto-detect your age and then asking for age-verification and plethora of other issues that a user is exposed to just by using the app which they might be using for the community itself, I really dislike it and I do wish if forums or something better could be used more as compared to discord.)
[Side-note: I built https://mirror.forum which is intended for communities to have both discord/fluxer & also other discord alternatives like stoat/matrix links to eventually help migrate users to these platforms and to help in finding the communities present in these spaces in the first place, I would love some feedback as the website is mostly crickets to be honest]
Fluxer.gg is really good and in my opinion, tries to be the most similar alternative to discord in terms of UI/UX so it should "just work"
I highly highly recommend you checking it out, I must say that the first time I used it, I was mind blown that this open source discord alternative actually feels so usable in day-to-day lives.
There are many alternatives to be honest, stoat, matrix,fluxer but fluxer is the alternative I would recommend for public spaces.
The issue now is I think mostly just moving the userbase over to there.
[I have talked to the dev and tried to share it a few times on hackernews as a submission to help get more userbase as I think that HN would definitely love fluxer if it came to know more about it]
(Very slight advertisement but I have tried to build something in this space at https://mirror.forum which can allow a community to have discord,fluxer,matrix,stoat links in the same place for the purposes of discovery of servers in other places than discord and also for the migration of the userbase itself as i think that the problem of an open source solution is mostly solved and this was my attempt at trying to partially help in fixing the moving the userbase issue and making people more familiar with alternatives, though the website mostly has no activity nowadays and I would love activity to happen but I am unsure how to realistically approach so... that's why I am advertising perhaps, I would love to get feedback as well :-D)
Crypto has been sold many many ways, but most of it is not a stable store of value.
Stablecoins are what are supposed to be stable sources of value, but Zeke Faux's book Number Go Up points out how opaque and unaccountable the companies managing these currencies are. From his reporting, their "tractor app" appears to be not so much farmers in sub-saharan Africa receiving aid payments but criminals looking for anonymous ways to transfer lots of money.
The blog post provides some context but no understanding.
It's meaningless because it's unverifiable. Dario all but said we wouldn't notice if they were pacing or not, because they have no intention of stopping development. The bits we could in theory verify are the external audits, whose independence has already been called into question.
At the very least, dumping a new model on the world before the ink on the glossy brochures of "Pacing the Frontier" was even dry calls into question their commitment.
... in fact the "pace" seems to have all but increased AND the surface area of the "frontier" has all but increased: There's the Pareto frontier, the AGI frontier, the price frontier, the structured output frontier, the open weight frontier, the Chinese chip-trained frontier, the inference frontier, harnesses ...
Consumers are worth less than creators in the cosmic scheme of things. I believe we should be much friendlier to consumers than our laws currently allow, but there should be a balance, and that balance should not privilege consumers above all.
If it allows you to distinguish readily between human intelligence and computer intelligence, it seems quite relevant to the question of whether computers have achieved something akin to human intelligence.
It may not be useful for anything else, but at least it can say that.
But it's not relevant as a metric to gauge distance to human intelligence. Humans see individual letters, LLMs do not. If I asked you the relative activation of the cones in your retina as I showed you some solid color image, you couldn't do it. You simply do not have cognitive access to that information. But that says nothing about your intelligence.
A more accurate test would be to give it a list of words (or anything represented as a single token) and ask it how many times that token appeared. I'm sure they have no trouble at that task.
Yes, they have some alien failure modes. But that should be expect, they are an alien intelligence. I might be willing to grant that a lack of ability to reflect on its own level of knowledge is a demerit to it being generally intelligent. But then again it is largely an artifact of training. I suspect if there were a guessing penalty during pretraining they would develop or more readily communicate the strength/reliability of their knowledge.
I am not enthusiastic about criteria for human-like intelligence that imply that dyslexic people don't have human-like intelligence.
[EDITED to add:] I actually don't know whether dyslexic people find it difficult to count letters in words, if they have them already written down by someone else. I suspect they find it harder than people who aren't dyslexic. But perhaps "blind people whose spelling is poor" would have been better; I would not want to deny them human-like intelligence either.
I have slight dyslexia. I can't automatically write double consonants all the time.
But counting certain type of letters is very simple algorithmic task especially in written text. Any reasonably intelligent actually thinking thing should come up with algo and then execute it. Which to me sounds like reasonable minimum bar for general intelligence.
Exactly. It's got nothing to do with sensory input and everything to do with reasoning.
If someone asked me how many f's are in a word I hadn't seen before verbally, then a reasoned response would be that I don't know, but I estimate based on the syllables...or ask them to spell it out.
These are all the sorts of questions where general problem solving works, even if the conclusion is "I don't have enough data to speculate".
So that these models fall apart on it so readily means we're either grossly handicapping then with the requirement to "be helpful" or they just fail to recognize the problem and are just stochastically spitting out a high probability token sequence for the input.
If you aren't an excellent speller and you are asked to count the number of some letter occurring in a passage of text, you will look at its written/printed form and go through the letters one by one. This is, indeed, a pretty easy task and you will probably get it right if you're careful.
The models don't get to see the text written down. By the time your input reaches them at all it's been converted into tokens. By the time they start thinking about it its been converted into embeddings in a sort of concepts-and-word-fragments space.
I do think it's a definite weakness of most LLM systems that they are bad at admitting (maybe because they're bad at knowing) when they don't really know something. (I have the impression that Anthropic's models are better at this than OpenAI's, but that isn't based on careful research or anything.)
What's the actual behaviour of today's LLM systems on these questions when they're allowed to "think"? Someone upthread mentioned that Sonnet 5 at "medium" thinking level mostly gets them right but makes mistakes sometimes. It would be interesting if we could see what its chain-of-thought looks like in these cases.
... I just tried six questions of this kind on Sonnet 5 at "medium" thinking -- this is the default thing you get from free-Claude -- and it got them all right in a way that at least superficially looks as if it's spelling them out and counting. Obviously this isn't enough to guarantee that there isn't anything grossly wrong with its reasoning capabilities in this area, but it doesn't look to me like "falling apart" and it doesn't look like strong evidence that thinking of it as a stochastic parrot is helpful here.
> computers could count the Rs in strawberry since vacuum tubes. that measure is irrelevant.
I don't think it's irrelevant but perhaps not in the way you're assuming. When assessing AGI I'm not evaluating counting characters or even the execution of math operators at any scale or speed. As you observe, computer software from Regex to spreadsheets and Mathematica already handle that well. But AGI isn't about what computers can do, it's about whether AIs can do the specific things which, until now, have been uniquely human capabilities. Like understanding nuanced context and then coming up with novel approaches to solve a new kind of problem not relying on any specific prior training or knowledge (the 'G' is for General).
Most definitions of AGI start from a baseline that already assumes easily passing a Turing test and doing anything via text response that a high school graduate could. I ding LLMs not for failing to count but for failing to intuitively understand the nuanced context of a simple class of problem it hasn't seen in its training data. I fully understand that the reason LLMs fail letter counting is that they operate at the token level. They weren't trained on individual letters first, like human 2nd graders.
The only reason recent LLMs get strawberry and blueberry correct now is that they have those words on their pre-training 'cheat sheet'. However, the underlying fundamental weakness in the way LLM intelligence works which leads to this failure mode still hasn't been addressed. Even when the frontier labs add "recognize any sub-token counting question and write a Python script" to the training cheat sheet so LLMs always pass that test... they'll still be unable to recognize a simple class of problem which isn't on their 'cheat sheet'. As long as that's the case, to me, they aren't AGI because they can't fully replicate human-like recognition of novel problem classes. And it's not just about letter-counting. That gap and others like it lead to many other kinds of non-human brittleness in LLM problem solving. Those are the classes of reasoning, intuition and insight that the ARC-AGI series has been trying to queue up as targets. Not to show how bad LLMs are but to help them be great in all these counter-intuitive edge cases
> planes don't flap wings therefore they cannot fly?
This example still misses my point, which isn't related to usefulness or economic value. I concede that LLMs can have greater utility and economic value than humans on many tasks. The point is most definitions of AGI include something like "can fully replicate all the routine daily tasks done by any competent high-school graduate." That's not related to whether LLMs can solve many high-value problems faster and at larger scale than any human. That was also true of ENIAC in 1946.
The fact an airplane can fly faster and farther than any bird is irrelevant to whether an airplane can "fully replicate all the routine daily tasks done by any competent bird." That's the bird equivalent to most AGI definitions. An airplane can't build a nest or recognize the signals encoded in birdsong.
In the same way airplanes fail the 'bird replacement' requirement, AIs currently fail most AGI requirements only on the terms: "fully", "all" and "any". And in this context, airplanes scoring 15,000% more than birds on 'speed' and 'distance' doesn't matter any more than AIs scoring 15,000% more than humans on 'add 10,000 numbers'. We still aren't near AGI because LLMs cannot fully match any high-schooler's ability to independently conceive new approaches to novel problems not in their prior training data.
which gets us closer to philosophical questions which I'm personally not that interested in.
>In the same way airplanes fail the 'bird replacement' requirement, AIs currently fail most AGI requirements only on the terms: "fully", "all" and "any".
I'm not sure we want a machine that fully succeeds that test.
Planes pass the 'bird replacement' test on the only criteria that matters to us ... flying.
If we wanted nest making planes I think we'd have them by now. Nest making doesn't rate highly on the problems we're looking to solve though.
I don't want a machine that is moody, or depressed or has schizophrenia, which are all pat of the human condition.
We don't need the human "intuition magic dust" to do 99.99999% of useful work.
They're machines designed to do the work we don't want to. That's as "general" as their intelligence needs to be.
I'd prefer if my clothes folding machine did not have an existential crisis.
>I'm confused, is your argument something like "It's too easy so AGI doesn't need to be able to do it"?
Do you possess magnetoreception? a stupid pigeon can "see" the earth's magentic field. why are you blind to it? does a lack of magnetoreception make your intelligence any less "general"
no, you're just blind to it because that's just the way it is.
LLMs are blind to character counting because that's the way they are.
It didn't stop ChatGPT from finding the Jacobian Conjecture counterexample.
Human intelligence and machine intelligence are only going to cross over to a certain degree.
same as plane flight and bird flight are only kinda related.
It matters under the lens of AGI. Artificial intelligence that can meet or exceed human intelligence. Sure, it's good at some things but rather limited at others.
Breadth of capabilities matters... and a promotional video is nice and all, but people are throwing this term around like it's a prize they've won, but they've not gotten there yet.
Maybe wages if the company puts the IPO proceeds towards additional hiring. The other bits sound pretty speculative to me. I could see outside investors walking away with most of the early gains, leaving employees and locals underwater.
Then they aren’t affecting housing. If gains are being realized to convert into real estate, they’re getting taxed. If they aren’t getting taxed, they aren’t hitting housing. (There are edge cases. But this isn’t an edge-case hypothesis. The folks selling shares in the private markets to buy mansions are paying taxes.)
It is not OpenAI's or Anthropic's responsibility to build housing in SF. It is the city's. Putting the heat on the mayor and supervisors is the correct move.
I've been "entertaining" myself recently with a report on how various Bay Area cities are tracking against their housing elements. Remember, these housing elements were their state-mandated plans for additional housing units between 2022 and 2030, and extra emphasis was put on this cycle to make sure the plans were realistic, accountable, addressed the magnitude of the need, and included (relatively) affordable housing.
We are close to halfway through this time period. San Francisco has built, according to this report, less than 10% of their commitment. They have a little over four years now to come up with 76K units. See: https://cities.fairhousingelements.org/cities/san-francisco
Last year SF built 1.75K units. This year, by June, they had built 400 units.
The federal government should step in. Give San Francisco four years to build the 76,000+ homes it needs. If the city fails, impose a substantial penalty for every unit missed and authorize the federal government to take over suitable public land and build the housing itself. At some point, housing shortages require consequences.
I don't think the local political mess in San Francisco warrants breaking two hundred years of federalism. Nothing does, really.
Also worth keep in mind the US tried mass public housing projects in the 60s. And then they stopped because they all ended up being crime ridden shitholes.
Those were arguably due to neglect rather than public housing being problematic per se. Perhaps the same thing would happen if the Feds tried it again; perhaps the neglect is a practically unavoidable consequence given our politics. But I think it should still be on the table. We tilted at the Bank of the US a few times before figuring out the Federal Reserve, after all.
But I agree, it would be far preferable to let property developers solve this problem. I support regional and state-level interventions and accountability. I don't want this particular federal administration within a thousand miles of the problem!
Carrots work better. Bankroll HSR and the burying of PG&E’s power lines in exchange for Sacramento forcing San Francisco to build. Tie funding distribution to actual housing starts.
Anthropic and OpenAI don’t have that many employees compared to most big employers. Their employees are highly compensated. The ratio between the housing the occupy and the taxes paid by their employees has to be one of the most favorable ratios of all companies in the area.
Losing those companies and their employees would cost the cities far more than they might gain.
Makes sense financially, but I suspect that the reason this isn’t all that appealing politically is that ordinary people can’t get those jobs. “We have a few, very rich employees” doesn’t seem like much of a selling point. It’s another example of inequality.
Data center finances seem vaguely similar? They could be a good source of tax revenue but don’t provide many jobs.
Silicon Valley exists as a persistent thumb on the nose against load-balancing ideas. It's had an absolute glut of tech companies for many decades now, with no apologies needed. Many people in the region benefit from that concentration of talent, wealth, and capital.
It does seem that way nowadays, but I wonder why more tech companies don't bring it to a different city, like Microsoft did in Redmond. The old tech companies did things like that but the newer companies mostly aren't for some reason.
I suppose there are some moving to Texas. But the East Bay is a lot closer.
The East Bay does have some spillover! But there are various reasons it's not as attractive. My hunch is that one of the major reasons is that commutes from the Peninsula and Silicon Valley itself are rough.
I'm not entirely sure they will, though.
reply