titaniumrain

If this problem has existed before, why start worrying now? And if scale might make it problematic, can we quantify the impact instead of simply worrying?

900 downloads in 5 days. not a very good sign, believe it or not

seems a DOA

These tips seem like common knowledge — essentially, it’s similar to a reminder to be cautious about what you eat when you haven’t prepared the meal yourself.

The author is the only one knows something about AI! wink wink wink

This is hardly news TBH. Check this paper out

https://www.ndss-symposium.org/ndss-paper/understanding-worl...

LLMs aren’t suited for everyone or every task—at least not yet. However, that won’t prevent their adoption in areas where they excel, such as code generation, translation, and similar domains.

history repeats itself - "horses are more efficient than cars" In addition, a study based on 16 devs is representative enough to draw this conclusion?

This post is hilarious. People like this author are the ones vetting start-ups? Please. The idea that alignment leads to a degradation in model utility is hardly news.

But let’s be clear: fine-tuning an LLM to specialize in a task isn’t just about minimizing utility loss. It’s about trade-offs. You have to weigh what you gain against what you lose.

“If I had asked people what they wanted, they would have said faster horses.” - HENRY FORD

The world moves on. Dinosaurs either die or evolve.

4,5 years as a PhD student, then 2,3 years as a postdoc then 5 more years to get tenured... modern slavery isn't it?