One of my grandparents had AD and became strangely lucid one day, a couple weeks before her death. It was like she finally woke from a years long fog.
I remember it vividly because I got to tell her I love her and she smiled and said it back to me. She seemed to understand me for the first time in years.
It was short lived but I'm very grateful for that single exchange it gave me.
It's not impossible to comply with. It will just take additional effort to ensure compliance.
For example, they can put this burden on enterprise customers to verify and attest citizenship. This is commonly done today for some types of cleared work.
For consumers, I'm sure it can be done if the monetary incentive is there. People will hate it, but it can be done.
Assuming it was cleverly designed to be impossible to comply with is giving far too much credit.
Back with some papers. (Apologies in advance; I typically don't edit/format comments much here, please bear with me.)
Notable papers describing performance improvements with prescribed roles and personas:
- ExpertPrompting: Instructing Large Language Models to be Distinguished Experts (2023) https://arxiv.org/abs/2305.14688 (if you're going to only read one paper here, maybe read this one but know there has been a lot of follow up with more modern models.)
A TLDR of my *admittedly heavily biased* mental model (so take it with a grain of salt): personas do improve task alignment and precision to measurable effect but with observed negative impact to accuracy and knowledge grounding. Overall, this makes it quite suitable and preferred for code generation scenarios. (Don't over-index on 'accuracy' here as meaning "bad code", it's more about verbosity/jargon reducing clarity of higher order goals like business objectives and system architecture.)
Outside of code generation, personas have the interesting effect of increasing implicit biases and stereotypes. It's not hard to imagine something like "you are a left|right wing politician ..." or "you are a senior-citizen|teenager ..." influencing token space construction considerably.
> Was that ever actually shown to be effective? Is it still?
Yes! Personas demonstrated measurable improvement in a few different ways, with caveats of course. The common intuition is that personas influence token space in beneficial ways.
I'll come back here later on desktop and link a few (still) relevant papers on this topic.
I remember it vividly because I got to tell her I love her and she smiled and said it back to me. She seemed to understand me for the first time in years.
It was short lived but I'm very grateful for that single exchange it gave me.
reply