Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Saying Goodbye to GitHub

ersei.net

Post reply on HN
Print view View original

421–430 of 450 posts

Re: Saying Goodbye to GitHub

#421

Post by m4nu3l »

usrbinbash wrote: post #394
m4nu3l wrote: post #314

Earlier quoted context omitted.

> But being better at mimicking reason, is still not reasoning How do I know people are not using a similar process when they perform "reasoning" but with a way more elaborate model? Can you prove me that the two are inherently different in the type of output they produce regardless of how large a ML model is or can be? Because if you can't, and they produce the same type of output, the processing could be similar en…

> but with a way more elaborate model? Simple: I know that humans have intentionality and agency. They want things, they have goals both immediate and long term. Their replies are based not just on the context of their experiences and the conversation but their emotional and physical state, and the applicability of their reply to their goals. And they are capable of coming up with reasoning about topics for which the…

> I know that humans have intentionality and agency.

You assume that. You can only maybe know that about yourself. But my question was bit different. How do you know that the ML model doesn't?

> about topics for which they have no prior information, by applying reasonable similarities.

This is a contradiction. If you have no prior information about a topic you can't know even what topic is similar.

> Even if someone never heard the phrase "walking a mile in someone elses shoes".

Same for ML modes. They don't have a representation of every possible prompt.

Re: Saying Goodbye to GitHub

#422

Post by pms »

bruce511 wrote: post #151
dijit wrote: post #74

Earlier quoted context omitted.

> I'm not aware of any Open Source license,or Free license for that matter,that has a give-back clause. §5.c of the GPL Version 3 states explicitly: > You must license the entire work, as a whole, under this License to anyone who comes into possession of a copy. This License will therefore apply, along with any applicable section 7 additional terms, to the whole of the work, and all its parts, regardless of how they…

>> As in, all modifications must be made available. Is that not meeting your definition of giving back? Available to all users. Not previous authors. There may be overlap, or there may not be overlap. Plus, I would say it's giving forward, not back. If there are public users then the original authors can become users and get the code. But there will be bug fixes and features smooshed together. Which is why i posit th…

Wow! Just wow... Apparently, some people don't get the idea of common good... as if it didn't exist...

Re: Saying Goodbye to GitHub

#424

Post by lofaszvanitt »

lelanthran wrote: post #384
lofaszvanitt wrote: post #298

Earlier quoted context omitted.

If there would be no open source, people would pay for libraries. Now we have open source, and a lot of devs are not compensated. End of story. No proper solution. That's all. Roughly the same applies to newspapers. Ohm please do not turn off advertisements so we could keep the lights going. Digital beggars everywhere.

> If there would be no open source, people would pay for libraries. Nonsense. The cost of creating non-trivial software (say, 20+ dependencies, all needing payment) would put software out of the reach of ordinary people, meaning that there will only be a small niche of developer jobs. Which means that most people making a non-zero income from writing software today would have been making a zero income from writing so…

I think you have issues with interpreting the idea as a whole, so you cling to one sentence and base some totally out of touch assumption on that very sentence.

Re: Saying Goodbye to GitHub

#425

Post by _gabe_ »

usrbinbash wrote: post #413
_gabe_ wrote: post #310

Earlier quoted context omitted.

> but the fact that they can also spew such complete illogical nonsense shows that they are not "reasoning" about things Have you ever seen the proof that 2=1 ? It looks convincing, but it's illogical because it has a subtle flaw. Are the people who can't spot the flaw just " looking like they are reasoning", but really they just lack the ability to reason? Are witnesses who unintentionally make up memories in court…

> Are the people who can't spot the flaw just "looking like they are reasoning", but really they just lack the ability to reason? Lacking relevant information or insight into a topic, isn't the same as lacking the ability to reason. > You can't just spout that an LLM lacks reasoning without first strictly defining what it means to reason. Perfectly worded definition available on Wikipedia: Reason is the capacity of c…

You must have missed the part where I said:

> Until we can come up with hard metrics that define these terms, nobody is correct when they spout their own nonsense that somehow proves the LLM doesn't fit into their specific definition of fill in the blank.

"Consciously", "logic", and "seeking the truth" are not objectively verifiable metrics of any kind.

I'll repeat what I said: Until we come up with hard metrics that define these terms, nobody can be correct. I'll take investopedia's definition for what a metric means, as that embodies the idea I was getting at the most succinctly:

> Metrics are measures of quantitative assessment commonly used for assessing, comparing, and tracking performance or production.[0]

So, until we can quantitatively assess how an LLM performs compared to a human in "consciousness", "logic", and "seeking the truth", whatever ambiguous definition you throw out there will not confirm or deny whether an LLM embodies these traits as opposed to a human embodying these traits.

[0]: https://www.investopedia.com/terms/m/metrics.asp

Re: Saying Goodbye to GitHub

#426

Post by fariszr »

galangalalgol wrote: post #284
fariszr wrote: post #33

> The code that was regurgitated by the model is marketed as "AI generated" and available for use for any project you want. Including proprietary ones. It's laundering open-source code. All of the decades of knowledge and uncountable hours of work is being, well, stolen. There is nothing being given back. Leaving GitHub wont change that, OpenAI is training its models on every bit of code they can have, sourcehut, cod…

I'm mostly familiar with gitlab, what does github provide for free above and beyond that? I like that I can run my gitlab pipeline on my machines and sync to a free gitlab instance. I like that I don't read about security vulnerabilities in gitlab pipelines nearly as often as github actions. I like gitlab issues as they are fairly minimal.

GitHub registry, GitHub actions and GitHub Codespaces are unlimited for public repos, in addition to all enterprise features.

That's without talking about nice to have features like GitHub Sponsors, the for you tab, the (arguably) more popular UI layout, It's simply a better platform for Open source projects

Re: Saying Goodbye to GitHub

#427

Post by fariszr »

rascul wrote: post #403
fariszr wrote: post #33

> The code that was regurgitated by the model is marketed as "AI generated" and available for use for any project you want. Including proprietary ones. It's laundering open-source code. All of the decades of knowledge and uncountable hours of work is being, well, stolen. There is nothing being given back. Leaving GitHub wont change that, OpenAI is training its models on every bit of code they can have, sourcehut, cod…

> gitlab doesn't even compare in free features. What features is GitLab missing? I don't know, I'm curious.

Unlimited package registry, unlimited Action run time, premium features unlocked and more. Also, the free tier on GitHub gives more for private repos too!, unlimited orgs, 2000 Ci minutes etc. It's just plain better, and It's because Microsoft can afford to play the long game, GitLab can't anymore.

Re: Saying Goodbye to GitHub

#428

Post by fariszr »

toastal wrote: post #190
mattlutze wrote: post #147

Earlier quoted context omitted.

Parent is talking about a fundamental feature of networks. A denser and larger network has much more useful network-related features, and if one company has a significant majority of the total addressable market for a network, it's a massive ask for people to extricate themselves and rebuild a network somewhere else. It's why Facebook is still on top even though everyone hated it for a while; YouTube is the *only vid…

But we are developers, not my grandma. We ‘know better’ but haven’t been doing enough about it.

You are overestimating how many developers care about this really.

Re: Saying Goodbye to GitHub

#429

Post by FeepingCreature »

usrbinbash wrote: post #416
FeepingCreature wrote: post #415

Earlier quoted context omitted.

But the human mental model is purely internal. For that matter, there is strong evidence that LLMs generate mental models internally. [1] Our interface to motor actions is not dissimilar to a token predictor. > The mental model we build and update is not just based on a linear stream, but many parallel and even contradictory sensory inputs So just like multimodal language models, for instance GPT-4? > as experiences…

> For that matter, there is strong evidence that LLMs generate mental models internally. Limited models, such as those representing the state of a game that it was trained to do: Yes. This is how we hope deep learning systems work in general. But I am not talking about limited models. I am talking about ad-hoc models, built from ingesting the context and semantic meaning of a string of tokens, that can simulate reali…

The relevant keyword you want is "zero-shot learning". (EDIT: Correction; "in-context learning". Sorry for that.) LLMs can pick up patterns from the context window purely at evaluation time using dynamic reinforcement learning. (This is one of those capabilities models seem to just pick up naturally at sufficient scale.) Those patterns are ephemeral and not persisted to memory, which I agree makes LLMs less general than humans, but that seems a weak objection to hang a fundamental difference in kind on.

edit: Correction: I can't find a source for my claim that the model specifically picks up reinforcement learning across its context as the algo that it uses to do ICL. I could have sworn I read that somewhere. Will edit a source in if I find it.

edit: Though I did find this very cool paper https://arxiv.org/abs/2210.05675 that shows that it's specifically training on language that makes LLMs try to work out abstract rules for in-context learning.

edit: https://arxiv.org/abs/2303.07971 isn't the paper I meant, since it only came out recently, but it has a good index of related literature and does a very clear analysis of ICL, demonstrating that models don't just learn rules at runtime but learn "extract structure from context and complete the pattern" as a composable meta-rule.

edit: I think I was thinking of https://arxiv.org/abs/2212.10559 , which asserts that ICL acts equivalent to gradient descent.

> In regard to my example given elsewhere in this HN thread: I know that Mike exits the elevator first because I build a mental model of what the tokens in the question represent. I can draw conclusions from that model, including new conclusions whos token-representation would be unlikely in the LLMs model, which doesn't explain anything about reality, but explains how tokens are usually ordered in the training set.

I mean. Nobody has unmediated access to reality. The LLM doesn't, but neither do you.

In the hypothetical, the token in your brain that represents "Mike" is ultimately built from photons hitting your retina, which is not a fundamentally different thing from text tokens. Text tokens are "more abstracted", sure, but every model a general intelligence builds is abstraction based on circumstantial evidence. Doesn't matter if it's human or LLM, we spend our lives in Plato's cave all the same.

Re: Saying Goodbye to GitHub

#430

Post by usrbinbash »

FeepingCreature wrote: post #429
usrbinbash wrote: post #416

Earlier quoted context omitted.

> For that matter, there is strong evidence that LLMs generate mental models internally. Limited models, such as those representing the state of a game that it was trained to do: Yes. This is how we hope deep learning systems work in general. But I am not talking about limited models. I am talking about ad-hoc models, built from ingesting the context and semantic meaning of a string of tokens, that can simulate reali…

The relevant keyword you want is "zero-shot learning". (EDIT: Correction; "in-context learning". Sorry for that.) LLMs can pick up patterns from the context window purely at evaluation time using dynamic reinforcement learning. (This is one of those capabilities models seem to just pick up naturally at sufficient scale.) Those patterns are ephemeral and not persisted to memory, which I agree makes LLMs less general t…

> In the hypothetical, the token in your brain that represents "Mike"

Mike isn't represented by a token. "Mike" is a word I interpret into an abstract meaning in an ad-hoc created, and later updated or discarded model of a situation in which exist only the elevator, some abstract structure around it, and the laws of physics as I know them from knowledge and experience.

> built from photons hitting your retina, which is not a fundamentally different thing from text tokens.

The difference is not in how sensory input is gathered. The difference is in what that input represents. For the LLM the token represents...the token. That's it. There is nothing else. The token exists for its own sake, and has no information other than itself. It isn't something from which an abstract concept is built, it IS the concept.

As a consequence, an language model doesn't understand whether statements are false or nonsensical. It can say that a sequence is statistically less likely than another one, but that's it.

"Jenny leaves first" is less likely than "Mike leaves first".

But "Jenny leaves first" is probably more likely than "Mario stands on the Moon", which is more likely than "catfood dog parachute chimney cloud" which is more likely than "blob garglsnarp foobar tchoo tchoo", which in turn is probably more likely than "fdsba254hj m562534%($&)5623%$ 6zn 5)&/(6z3m z6%3w zhbu2563n z56".

To someone reaching the conclusion that Mike left the elevator first by drawing that conclusion from an abstract representation of the world, all these statements are equally wrong. To a language model, they are just points along a statistical gradient. So in a language models world a wrong statement can still somehow be "less wrong" than another wrong statement.

---

Bear in mind when I say all this, I don't mean to say (and I think I made that clear elsewhere in the thread) that this mimickry of reasoning isn't useful. It is, tremendously so. But I think it's valueable to research and understand the difference in mimicking reason by learning how tokens form reasonable sequences, and actual reasoning from abstracting the world into models that we can draw conclusions from.

Not in the least because I believe that this will be a key element in developing things closer to AGIs than the tools we have now.

Post reply on HN