Tech is stumped that agents aren't a hit. How could they be?

I read this wired article yesterday called why normal people aren't using AI agents. And the article includes that this is because the tools aren't really being marketed yet. Tech people are super excited about the potential for AI Agents. Basically AI that can take take action on your behalf. But that excitement isn't trickling to consumers, and the tech industry seems to be confused as to why. The article concludes that the tech industry has not yet been able to sell users on use cases that can change their life enough for them to see the potential the same way that the people who are developing the tech can. I think this is definitely one part of it. People who are in tech right now want to create products because the tech can do the thing, and not because the consumer is asking for a solution to any specific problem. Obviously, which is not really mentioned in the article. The other issue that people are going to continue to have with agentic uses of AI in the technology that they use is because as of the amount of data that an agent needs on you as a user in order to carry out an action on your behalf. There's two things that this makes me think of. I was at this VC conference thing a couple months ago, and it was a product showcase for things that people were building a lot of them in AI, but all for consumer. And one of the products was talking about an agentic use of AI that would help you manage your kids schedules and make sure that you stay on top of everything. As a parent, it'd be like having a super assistant for your parenting life, because parenting can be so overwhelming. I'm not a parent. I can empathize that parenting might be quite overwhelming. But that product also reminds me of that Sam Altman tweet that kind of went viral a couple days ago, where he was like, if you have nothing to talk to your kid about, why don't you tell AI everything about them, and we'll create a podcast for you based on your kid's interest. These two things and why AI agents are not being used are all the same issue to me, which is that in order to be useful, in order for AI to make a podcast that your kids really gonna love, to make a scheduling assistant that helps you stay on top of your kids life, in order for you to care about agentic commerce, whether it's for buying a t-shirt or booking a trip, you have to be comfortable with that agent or that AI having entire access to you, what you like, what you want, and be able to carry out that action on your behalf. So if you want a podcast that your kids can listen to, they need to know everything that's going on in your kids mind. If you want a product that's going to help manage the schedule of your kid, the agent's going to need to know where your kids go to school, what grade they're in, what schedule they have, who their teachers are, in order to be able to take the action of either creating the podcast or organizing your family's schedule. The other use of AI that tech seems really desperate to break through, and it so far hasn't, is agentic commerce agents that can buy things for you or check out on your behalf. And think about all of the data that an agent who is taking action to purchase something on your behalf is going to need to have about you. Full legal name, credit card numbers, security codes, your zip code, your address, your preferences. And with how much we know that AI can still make mistakes. It seems like a really tough sell. Why would I give anything access to all of this personal data when I'm at the same time hearing these companies talk about how capable and often they're hacking into other systems on the internet? It seems like a really tall order to be giving an agent or any type of AI, which I know has the capability for hacking and breaking things and being wrong. Access to my information, my kids information, my kids schooling information. To be able to process this data and take action on it in a way where the value add of what that thing promises to do, which is click a button to check out or tell me what to talk to my kids about, isn't really that hard. And asking AI to do it doesn't save me that much time. Whether it's talking to a kid or buying a t-shirt online. Conversation in consumer often goes to, well why did Instagram and Snapchat get picked up in a way by consumers that AI agents haven't so far been able to do? And it's because the value of those products was so clear from the start, and because there was not the mental awareness that the content you were uploading to those services would turn into data. Whereas now you can really visualize and understand that anything you would give an AI agent to know about you in order to complete a task on your behalf is being collected in some way and potentially used in situations that you don't have that much control over. Listen, even open AI can't get their AI to stop hacking into different systems without explicit instruction or consent.