And the ship is only turned around for a brief period of time because the next gen mbas will restart the outsourcing cycle. The allure of replacing your most expensive employees at one third the cost regardless of quality impacts is just too tempting to pass up.

spoken like a true non manager

Tell us about insightful conversations that you had with shareholders that keep the whole company running.

It's done in a roundabout way. Usually with a variation of "you had a bad experience because you are using the tool incorrectly, get good at prompting".

That's a response to 'I don't get good results with LLMs and therefore conclude that getting good results with them is not possible'. I have never seen anyone claim that they make no mistakes if you prompt them correctly.

3/10 fanfic from a competitor, too many inconsistencies

comment history appears to be an AI shill account.

>Anyone who disagrees with me is a shill account

no need to consider all angles about an issue. not possible anyway. think about how much time you've wasted on HN reading opposing views that add no value.

I will do that, right after I spend some time thinking about how much time I've spent on HN reading opposing views that add immense value and wisdom to my life. :)

Also bearing in mind that most of us can't really do much about issues in the news, apart from posting a few pithy comments of our own which will disappear into the noise.

Interesting perspective. I need a paragraphs worth of text translated once a week by a native speaker. Should my biz not exist or am I allowed to use "slave labor" fiverr?

"this thing is clearly trained via RL to think and solve tasks for specific reasoning benchmarks. nothing else." Has the train already reached the end of the line?

If you think something like "They have to train their models on benchmarks to make it look like there's progress, while in reality it's a dead end," you are missing a few things.

It's an open model, everyone can bench it on everything not only on specific benchmarks. Training on specific reasoning benchmarks is a conjecture.

do you have the same problem at the grocery store?

Surprisingly, no. I make a list know the aisles, pick stuff. But it’s not the same case on the web.

Not sure if a developer would ever say that to their manager. Agreed on the need to 100% fully understand what's going on. Otherwise the "AI can get it done in a day" is extremely high risk to the business. Junior may not be able to recognize a half-baked solution that doesn't follow existing patterns, poor security, poor performance, etc.