4 ms·

It's been used for defense for a long time. AI is why spam filters started actually working in the 2000s.

by asdfman123 3mo ago

PunchyHamster 3mo ago
different kind of AI
- kazinator 3mo ago
  Not really, though. Sequences of tokens are predicted as {spam, not spam} via training on pair sets. It's like a token predictor where the output vocabulary is two tokens.
  - Dylan16807 3mo ago
    The complete inability to use it to generate spam doesn't make it different to you?
    - kazinator 3mo ago
      Well, I just said that. Since what is predicated is one of two tokens <is-spam>, <is-not-spam>. Since spam is not made out of those tokens, it doesn't generate spam. (Even spam could be encoded into those two tokens via binary or whatever, it still wouldn't because it's predicting a classification, and not the next sybmol; i.e. not simply the next bit in a after a string prefix, but a 0 or 1 value indicating whether the bit string is spam, which comes from an asserted external judgment.) It's all the same sort of thing though: some functions trained to predict a value associated with an input.
      - Dylan16807 3mo ago
        You did say that. And then despite that you reached a conclusion that I don't see as compatible. There is a massive qualitative and quantitative gap between the two. Even if you had a real token predictor for the word spam, you would need a thousand of these to touch LLM capability. But it's not a real predictor. It's not locational. It's something vastly weaker. > It's all the same sort of thing though: some functions trained to predict a value associated with an input. Now you made it even worse. If not for the word 'trained' this would be every program. If two programs fall under the umbrella of machine learning in any way, that kind of logic treats them the same. What would you consider a "different kind of AI"? Your definition there fits basically the entire field.
spydum 3mo ago
That was mostly Bayesian probability filters, followed by SPF and more recently DKIM.. Maybe you could consider the Bayesian filter databases being trained as ML, but I feel it's a far way from our typical AI/LLM surge
- TremendousJudge 3mo ago
  less than 10 years ago I worked for an AI company that sold Naive Bayes classifiers as a service. so I would call it AI
- dosisking 3mo ago
  > but I feel it's a far way from our typical AI/LLM surge I think they are about the same. Both work about as well as autocorrect
- inigyou 3mo ago
  Yet another instance of when something from the field of AI actually works, it stops being called AI.

It's been used for defense for a long time. AI is why spam filters started actually working in the 2000s.

PunchyHamster 3mo ago

different kind of AI

kazinator 3mo ago
Not really, though. Sequences of tokens are predicted as {spam, not spam} via training on pair sets. It's like a token predictor where the output vocabulary is two tokens.
- Dylan16807 3mo ago
  The complete inability to use it to generate spam doesn't make it different to you?
  - kazinator 3mo ago
    Well, I just said that. Since what is predicated is one of two tokens <is-spam>, <is-not-spam>. Since spam is not made out of those tokens, it doesn't generate spam. (Even spam could be encoded into those two tokens via binary or whatever, it still wouldn't because it's predicting a classification, and not the next sybmol; i.e. not simply the next bit in a after a string prefix, but a 0 or 1 value indicating whether the bit string is spam, which comes from an asserted external judgment.) It's all the same sort of thing though: some functions trained to predict a value associated with an input.
    - Dylan16807 3mo ago
      You did say that. And then despite that you reached a conclusion that I don't see as compatible. There is a massive qualitative and quantitative gap between the two. Even if you had a real token predictor for the word spam, you would need a thousand of these to touch LLM capability. But it's not a real predictor. It's not locational. It's something vastly weaker. > It's all the same sort of thing though: some functions trained to predict a value associated with an input. Now you made it even worse. If not for the word 'trained' this would be every program. If two programs fall under the umbrella of machine learning in any way, that kind of logic treats them the same. What would you consider a "different kind of AI"? Your definition there fits basically the entire field.

spydum 3mo ago

That was mostly Bayesian probability filters, followed by SPF and more recently DKIM.. Maybe you could consider the Bayesian filter databases being trained as ML, but I feel it's a far way from our typical AI/LLM surge

TremendousJudge 3mo ago
less than 10 years ago I worked for an AI company that sold Naive Bayes classifiers as a service. so I would call it AI
dosisking 3mo ago
> but I feel it's a far way from our typical AI/LLM surge I think they are about the same. Both work about as well as autocorrect
inigyou 3mo ago
Yet another instance of when something from the field of AI actually works, it stops being called AI.