This is the real killer. Gambling in the CS scene is absolutely rampant, you will only find a handful of CS influencers that do no accept gambling sponsors (As they offer much more than other companies).
I was happy when Belgium banned loot boxes, but it was not nearly enough.
What's the deal with trying so hard to defend the advertising of an objectively harmful industry?
I think a good parallel here is Tobacco advertising. Smoking is harmful in every way, therefore I don't see any reason why advertising it to a broad audience (which will inadvertently also include children) is something we should allow.
What's the net benefit of allowing such advertising? I don't see it. Yo could argue something about rights to free speech or some variation, but societies still have a responsibility to look out for the health of their people, no?
So, because another party is disproportionately punished for their crimes, we must punish others disproportionately as well? We should be raising the bar, not giving it a funeral.
I don't think that's a good way of thinking about it. Just because programmers might have better job security doesn't mean other motivators for not wanting to get fired exist.
As someone who played a lot of CSGO as a 15-16 year old back in 2015, it's all psychological. Why does anyone become addicted to gambling? The thrill releases enormous amounts of dopamine. Satisfy this over and over, and the threshold for the same euphoria becomes higher and higher. That's how yo end up with people, and particularly teens and young adults spending massive amounts of money for 'some pixels'. I think I sank about 1.5k into it in total over the years. When Belgium decided to ban such loot boxes I was happy to see the addiction forcibly ended.
They just need to repeat it for predatory battlepasses and the likes.
Imo it depends on what you're prompting. If you yourself do not fully understand what you want and/or which information you have to provide for the model to 'understand' enough context, prompts will not give good results.
I usually find ChatGPT very good for converting conceptual ideas to a concrete piece of 'boilerplate' to work off of. It can help you jump gaps in your proces, but in the end you still have to make the jumps yourself.
We do natural language processing with some custom tokenisation, the data it uses at runtime is put into an sqlite db at build-time. Works fine for our usecase.