8 ms·

For me the LLM topic is just all enveloping in a way the other trends haven't been. Obviously there was a a million crypto posts and it was annoying but crypto

For me the LLM topic is just all enveloping in a way the other trends haven't been. Obviously there was a a million crypto posts and it was annoying but crypto covered a specific niche of overall software stuff (payment processing). With LLMs it feels like every topic somehow has to come back to LLMs. I currently work as a ML Eng at FAANG so maybe it adds to my exhaustion on the topic.

TheMongoose 2y ago
It's deeply funny that they hit you with the ole "you don't understand AI" when it's your day job though. So you've got that going for you.
incanus77 2y ago
> For me the LLM topic is just all enveloping in a way the other trends haven't been. Same here, for sure. I just try to dodge it all as best I can. Seems like every question has the possible answer of LLMs, and nearly always, someone provides it.
- somenameforme 2y ago
  There's also this annoying dissonance between 'evangelists' and reality. Evangelists often feign that we're on the cusp of artificial general intelligence, while in practice LLMs remain stupid, error prone, and unreliable for anything beyond tasks that have a vast number sources for highly relevant training. Which also somewhat dilutes the utility, because I could just as well find those sources! Oh right, that must just be because I'm not giving it the magic prompt that makes them magically turn into geniuses.
  - jcgrillo 2y ago
    Sometimes I wish I had the drugs that would make me believe in things like "we're on the cusp of AGI", everything would be so much more exciting.
  - astrange 2y ago
    The current trend in LLMs is synthetic data and inference-time reasoning. They're past data sources. The new problem is this only works when you can verify the answers.
re-thc 2y ago
> covered a specific niche of overall software stuff (payment processing) At the time it was everything blockchain. Not just payments. Decentralized, smart contracts and the like.
- 3vidence 2y ago
  It's true it did get pretty rediculous and somewhat collapsed (at least in public interest). I just don't see why every time we need to go through the incredible hype / death cycle. LLMs are a useful tool if used properly, I hope I start seeing them used to create distinct value.
  - bookofjoe 2y ago
    "Replicants are either a benefit or a hazard."—Deckard
seanmcdirmid 2y ago
Crypto currencies became purely speculative asses and their volatility ensured that they wouldn’t be useful for payment processing. Has block chain in general made any progress in payment processing?
- 3vidence 2y ago
  I very much agree, just that if you were talking about idk something like compilers people wouldn't barge in about crypto. Any HN topic now just feels like a small jump to LLMs. I recognize my above is a bit of a strawman, hard to recall exact posts / comments over the years.

TheMongoose 2y ago

It's deeply funny that they hit you with the ole "you don't understand AI" when it's your day job though. So you've got that going for you.

incanus77 2y ago

> For me the LLM topic is just all enveloping in a way the other trends haven't been. Same here, for sure. I just try to dodge it all as best I can. Seems like every question has the possible answer of LLMs, and nearly always, someone provides it.

somenameforme 2y ago
There's also this annoying dissonance between 'evangelists' and reality. Evangelists often feign that we're on the cusp of artificial general intelligence, while in practice LLMs remain stupid, error prone, and unreliable for anything beyond tasks that have a vast number sources for highly relevant training. Which also somewhat dilutes the utility, because I could just as well find those sources! Oh right, that must just be because I'm not giving it the magic prompt that makes them magically turn into geniuses.
- jcgrillo 2y ago
  Sometimes I wish I had the drugs that would make me believe in things like "we're on the cusp of AGI", everything would be so much more exciting.
- astrange 2y ago
  The current trend in LLMs is synthetic data and inference-time reasoning. They're past data sources. The new problem is this only works when you can verify the answers.

re-thc 2y ago

> covered a specific niche of overall software stuff (payment processing) At the time it was everything blockchain. Not just payments. Decentralized, smart contracts and the like.

3vidence 2y ago
It's true it did get pretty rediculous and somewhat collapsed (at least in public interest). I just don't see why every time we need to go through the incredible hype / death cycle. LLMs are a useful tool if used properly, I hope I start seeing them used to create distinct value.
- bookofjoe 2y ago
  "Replicants are either a benefit or a hazard."—Deckard

seanmcdirmid 2y ago

Crypto currencies became purely speculative asses and their volatility ensured that they wouldn’t be useful for payment processing. Has block chain in general made any progress in payment processing?

3vidence 2y ago
I very much agree, just that if you were talking about idk something like compilers people wouldn't barge in about crypto. Any HN topic now just feels like a small jump to LLMs. I recognize my above is a bit of a strawman, hard to recall exact posts / comments over the years.