AI shopping assistants are evolving fast, but there’s a growing question that feels important to ask now rather than later: who will these AI systems ultimately work for? Right now, most online platforms make money by selling user intent to advertisers.

With AI moving into commerce and companies experimenting with things like “BuyItInChatGPT”, are we heading toward a future where AI becomes even better at selling to us?

Is a different path possible? One where AI agents are aligned with people instead of platforms? What would need to change for that to happen: business models, data access, regulations, or something more fundamental? I wrote down some initial thoughts here.

Would love to hear perspectives from this community. What would it take to build an AI economy that users actually trust?

What if AI assistants didn’t belong to companies but to users?
byu/aeriefreyrie inFuturology

9 Comments

  1. or you can just spend 10k on a local setup and run a local open source model. Feel free to let other people use for free if you want.
    No need for some revolution crazy scheme, just spent 10k for a local model. Deepseek model is not that far behind, in programming and engineering, i think it’s actually better.
    If users cant afford, then group with other people to come up with that 10k. It really dependent on whether user view 10k is worth more than the privacy of their data.

  2. Luke_Cocksucker on

    I guess I’m confused about how AI works. Wouldn’t you need access to data centers and massive computing power to run these things? Or am I completely wrong?

  3. Remington_Underwood on

    A complete reversal of late stage capitalism and the world wide transition to a Roddenberrian social order.

  4. They will gather more data from us to feed it to their AIs so they can align their ads more precisely towards us.

  5. This is why I pay for my own subscription because I feel like my assistant is mine, not the company’s.

  6. I haven’t been to any company website where the AI chatbot was a useful addition. Literally. I’ve spent some time playing and working and building some Ai applications on side at work and it’s my belief that 99 percent of ‘Ai’ chat bots or implementations in companies are not needed.

    I called the rent company to be greeted with hi, I’m Alexa, your Ai assistant, please leave a message after the beep. LoL wtf

  7. No one is building a consumer/human friendly local agent that is yours instead of ‘theirs’ because the current market sees people not caring about giving up their info/lives to ad services in exchange for convenience. Companies make things that make money. We keep giving modern data companies the raw materials for free so they can sell ad-time to advertisers. These tools aren’t FOR US, they USE US to sell their product (ads). Until there is enough demand to commercialize a ‘personal server’ product that is a standard part of every home, that is easy to use, provides basic functions that are currently just ad-machines, and can host apps and math runable on consumer grade equipment easily, companies will keep the enshittification machine of our current era rolling. Hobbyists have proven it’s possible, but it’s complicated and no one is trying to make it consumer-practical yet.

  8. I used ai to help me buy a car. I only trusted it because I could verify the information it gave me with the manufacturer. No way I would trust a black box ai that just arbitrarily recommends products. I’d rather rely on astrology.

  9. KanedaSyndrome on

    You can already run local LLMs for whatever you need. Not as fast and requires a strong GPU