Comment by wnolens

2 days ago

Why would I want to spend 1-2h researching humidifiers if I can spend that time in any other way, and still end up with a humidifier that fits my needs first try?

This kind of task is perfect for AI in a way that doesn't take away too much from the human experience. I'll keep my art, but shopping can die off.

Because you end up with a $1 value piece of crap that someone spent considerable time optimizing and faking reviews for LLMs instead of on the product. Basically in the medium term this strategy will get you temu stuff

How often do you buy the first result on an Amazon search? Because that's delegating your labour, isn't it? Surely the best products are getting to the top, right? Well no, they're being paid to get to the top. An LLM that has in-app shopping is gonna be the same thing

> This kind of task is perfect for AI in a way that doesn't take away too much from the human experience.

Not the current form of AI. I regularly use Project Farm to find the best "insert tool". In an ideal world a robot runs all of these tests in perpetuity covering every physical appliance possible (with every variation, etc.). However, current AI cannot do this. Obviously LLMs can't do this because they don't operate in the physical world.

Well, you can always do the same thing that an LLM would: open SEO spam ranking sites "best humidifiers 2025", filled with referral links to Amazon or other sellers, which basically copy product descriptions and assign rankings that aren't based on any tests or real data.

for the same problems with amazon, youre relying on a computer to tell you what to buy, which is very shortly going to be infested with promoted products and adverts instead of genuine advice. The AI implementers will poison the responses in the name of advertising, of this i have zero doubt in my mind.

  • 10+ years ago you could in fact just pick the best-reviewed product on Amazon at a certain price point and have a great experience! God help you if you tried that today.