HNHacker News
TopNewBestAskShowJobs

Zylokloto

31 karma · joined August 23, 2026

submissionscomments
Zylokloto··on I were 17, I'd learn how to build LLMs from scratch
We finetune LLMs. Small ones like Gemma 4 for semantic tasks.

There are plenty of areas were we need people to do this for insurances, banks etc.

AI/ML exists on many levels.

Zylokloto··on Erik Brynjolfsson says an AI "job apocalypse" is unlikely
He clearly states that it worries him and he made a public letter.

The job disruption is real and it will be an job apocalypse for plenty of people.

If a company is now paying for ai tokens and you now have 5-20% higher cost but you are relevant more efficient, you will reduce your team by 5-20%.

It is already a job apocalypse for people who learned language and no longer get a job for translating or text analysis. People who want to be an illustrator have a job apocalpyse.

And we have not even SEEN what it will look like in 4-14 years. Physical AI / Robotic is progressing like crazy too.

Zylokloto··on Canada will match US tariffs 'dollar for dollar' as trade talks break down
You underestimate how much more investment is happing in EU to be more independent of the USA.

Im also pretty sure that people start looking more to China than before.

Zylokloto··on Why your local LLM feels dumber than it is
You could do this for sure but its quite a lot of effort and a normal software stack is quite well known due to the massive amount of github projects and other opensource code.

You can also get a lot out of a harness in this case or your Agents.md or Claude.md file by just enhancing the context.

You might even question yourself if you are doing something wrong if a modern LLM really struggles with your code.

We do the finetuning only on small semantic data were it helps a lot.

I'm still wondering when we see smaller models (faster and cheaper) for more specific stacks like spring boot + java + angular + english only or so. Interstingly enough, i assumed LLMs are really good in any language but it seems that non english languages do reduce the ooutput quality of an LLM. At least last years GTC there was a talk about it.

There are companies though which ahve this exact problem with programming languages you normally don't see. ABAP for example is a very well known SAP language.

← PreviousPage 2 of 2