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

Claude’s memory architecture is the opposite of ChatGPT’s

shloked.com

Post reply on HN
Print view View original

121–130 of 240 posts

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#121

Post by resters »

ankit219 wrote: post #37

The difference is implementation comes down to business goals more than anything. There is a clear directionality for ChatGPT. At some point they will monetize by ads and affiliate links. Their memory implementation is aimed at creating a user profile. Claude's memory implementation feels more oriented towards the long term goal of accessing abstractions and past interactions. It's very close to how humans access mem…

Suppose the user uses an LLM for topics a, b, and c quite often, and d, e and f less often. Suppose b, c, and f are topics that OpenAI could offer interruption ads (full screen, 30 seconds or longer commercials) and most users would sit through it and wait for the response.

All that is needed to do that is to analyze topics.

Now suppose that OpenAI can analyze 1000 chats and coding sessions and its algorithm determines that it can maximize revenue by leading the user to get a job at a specific company and then buy a car from another company. It could "accomplish" this via interruption ads or by modifying the quality or content of its responses to increase the chances of those outcomes happening.

While both of these are in some way plausible and dystopian, all it takes is DeepSeek running without ads and suddenly the bar for how good closed source LLMs have to be to get market share is astronomically higher.

In my view, LLMs will be like any good or service, users will pay for quality but differnet users will demand different levels of quality.

Advertising would seemingly undermine the credibility of the AI's answers, and so I think full screen interruption ads are the most likely outcome.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#122

Post by Eisenstein »

shakna wrote: post #54
techbruv wrote: post #22

Earlier quoted context omitted.

I don’t understand the argument “AI is just XYZ mechanism, therefore it cannot be intelligent”. Does the mechanism really disqualify it from intelligence if behaviorally, you cannot distinguish it from “real” intelligence? I’m not saying that LLMs have certainly surpassed the “cannot distinguish from real intelligence” threshold, but saying there’s not even a little bit of intelligence in a system that can solve more…

Scientifically, intelligence requires organizational complexity. And has for about a hundred years. That does actually disqualify some mechanisms from counting as intelligent, as the behaviour cannot reach that threshold. We might change the definition - science adapts to the evidence, but right now there are major hurdles to overcome before such mechanisms can be considered intelligent.

What is the scientific definition of intelligence? I assume that is it is comprehensive, internally consistent, and that it fits all of the things that are obviously intelligent and excludes the things which are obviously not intelligent. Of course being scientific I assume it is also falsifiable.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#123

Post by shloked »

extr wrote: post #13

They are changing the way memory works soon, too: https://x.com/btibor91/status/1965906564692541621 Edit: They apparently just announced this as well: https://www.anthropic.com/news/memory

Thanks for sharing this! Seems like I chose exactly the wrong day to write this

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#124

Post by shloked »

kiitos wrote: post #5

> Anthropic's more technical users inherently understand how LLMs work. good (if superficial) post in general, but on this point specifically, emphatically: no, they do not -- no shade, nobody does, at least not in any meaningful sense

I should've been clearer, but what I meant was language models 101. Normal people don't understand even basics like LLMs are stateless by default and need to be given external information to "remember" things about you. Or, what is a system prompt.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#125

Post by cantor_S_drug »

GuB-42 wrote: post #106
Workaccount2 wrote: post #48

Earlier quoted context omitted.

Don't fool yourself into thinking Anthropic won't be serving up personalized ads too.

Anthropic seems to want to make you buy a subscription, not show you ads. ChatGPT seems to be more popular to those who don't want to pay, and they are therefore more likely to rely on ads.

They might be coming from different directions. But these things, as often they do, will converge. Too big of a market to leave.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#126

Post by mrheosuper »

GuB-42 wrote: post #106
Workaccount2 wrote: post #48

Earlier quoted context omitted.

Don't fool yourself into thinking Anthropic won't be serving up personalized ads too.

Anthropic seems to want to make you buy a subscription, not show you ads. ChatGPT seems to be more popular to those who don't want to pay, and they are therefore more likely to rely on ads.

so ChatGPT will become "saleman". And i do not trust any saleman.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#127

Post by Eisenstein »

astrange wrote: post #84
codedokode wrote: post #35

Earlier quoted context omitted.

You don't want an AGI. How do you make it obey?

How do you make your own children obey? (Meta-question: since they don't do this, why does it turn out not to be a problem?)

I think you have proven their point?

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#128

Post by skissane »

lupusreal wrote: post #50
techbruv wrote: post #22

Earlier quoted context omitted.

I don’t understand the argument “AI is just XYZ mechanism, therefore it cannot be intelligent”. Does the mechanism really disqualify it from intelligence if behaviorally, you cannot distinguish it from “real” intelligence? I’m not saying that LLMs have certainly surpassed the “cannot distinguish from real intelligence” threshold, but saying there’s not even a little bit of intelligence in a system that can solve more…

What it really boils down to is "the machine doesn't have a soul" . Just an unfalsifiable and ultimately meaningless objection.

Maybe panpsychism is true and the machine actually does have a soul, because all machines have souls, even your lawnmower. But possibly the soul of a machine running a frontier AI is a bit closer to a human soul than your lawnmower’s soul is.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#129

Post by gitremote »

lupusreal wrote: post #50
techbruv wrote: post #22

Earlier quoted context omitted.

I don’t understand the argument “AI is just XYZ mechanism, therefore it cannot be intelligent”. Does the mechanism really disqualify it from intelligence if behaviorally, you cannot distinguish it from “real” intelligence? I’m not saying that LLMs have certainly surpassed the “cannot distinguish from real intelligence” threshold, but saying there’s not even a little bit of intelligence in a system that can solve more…

What it really boils down to is "the machine doesn't have a soul" . Just an unfalsifiable and ultimately meaningless objection.

Incorrect. Vertebrate animal brains update their neural connections when interacting with the environment. LLMs don't do that. Their model weights are frozen for every release.

Re: Claude’s memory architecture is the opposite of ChatGPT’s

#130

Post by furyofantares »

cj wrote: post #95
fluidcruft wrote: post #91

Earlier quoted context omitted.

But aren't we only worth something like $300/year each to Meta in terms of ads? I remember someone arguing something like that when the TikTok ban was being passed into law... essentially the argument was that TikTok was "dumping" engagement at far below market value (at something like $60/year) to damage American companies. That was something the argument I remember anyway.

If that’s the case, we have an even bigger problem on our hands. How will these companies ever be profitable? If we’re already paying $20/mo and they’re operating at a loss, what’s the next move (assuming we’re only worth an extra $300/yr with ads?) The math doesn’t add up, unless we stop training new models and degrade the ones currently in production, or have some compute breakthrough that makes hardware + operatin…

> If we’re already paying $20/mo and they’re operating at a loss

I'm quite confident they're not operating at a loss on those subscriptions.

Post reply on HN