alt Hacker NewsAgree with this article, and I almost threw up in my mouth when I read this quote from Stephen Levy:
> By the end of this decade, it’s unlikely that people will swipe on their phones to tap on Uber or Lyft. They will just tell their always-on AI agent to get them home. Or that agent will have already figured out where they need to go, and the car will be waiting without the friction of a request. “There’s an app for that,” may be replaced by “Let the agent do that.”
Who TF are these people who think this kind of future is desirable? I basically think it's just people that want to broadcast that they're so important and busy that they can't take the 5 seconds it takes to hail an Uber. Its like all that "productivity optimization" porn that people spew online to show how focused they are.
I was reading article recently that said that a majority of people interviewed did not want to use AI agents simply because they didn't have much stuff in their life worth automating. Or more to the point, a lot of people actually enjoy making grocery lists, planning trips, picking out gifts for friends, etc. This stuff is generally considered "life", not some back breaking drudgery like washing clothes in a stream that I'd like to automate.
These folks like Levy who view this dystopian future as some sort of nirvana (and not because they view a different future, they actually want all this nonsense) can go F themselves. You can also tell how incredibly sheltered these people are because you can see they're rarely interacting with people outside their bubble. For example, a lot of people that open the Uber app make their decision based on data in the app, like "surge pricing, nevermind, I'll just walk" or "this looks expensive, let me try Lyft". You could argue an agent could learn all those rules, but again, these minutia of life are not exactly a nuisance to most people.
Replies
That's just the experience any executive with a secretary or personal assistant is used to having.
If AI allows more people to have such a premium experience, that's a use of technology that makes a lot more sense than all the "AI will take over your job" scaremongering.
I'd be more inclined to believe that an abundance of robotaxis will use predictive algorithms to preemptively show up wherever they're likely to be needed, allowing a UX where users can hail them like traditional taxis without an app. Maybe not in four years, but maybe in a decade or two.
That feels both more credible and more desirable than the magic panopticon predicted in the quote, and doesn't really depend on any major technological leaps beyond continued maturation and scaling of Waymo/alternatives.
> Who TF are these people who think this kind of future is desirable?
Some of this is weird techno delusion. Some of it is because the people describing it do a poor job of explaining how it might work.
If a couple decades ago someone told you that you’d have an always listening device in your pocket to answer your questions from all the world’s information, it would have sounded dumb, and with the always-listening device, rather dystopian. But that’s what you have assuming you have any modern smart phone.
The “agent knows where you’re going and calls a car for you” sounds dystopian as hell if done totally autonomously. But you could also imagine that an agent pops up a message on your watch “hey, you’ve been at dinner for an hour, if you’re winding down I can call you a car in 15 minutes” and suddenly it’s not that absurd.
Why are you crashing out about this? It's just an example of how AI agents can be a better user experience.
> [...] broadcast that they're so important and busy that they can't take the 5 seconds [...]
It takes a lot more than 5 seconds to make an informed decision these days. Apps and websites are throwing abusive fine print and dark patterns at users left and right.
I'd be absolutely thrilled to e.g. not have to interact with the Uber app and all its dark patterns if there were somebody or something I could trust to competently represent my interests.
That said, that's a big if, i.e., whether commercial LLMs or agents will be able to do that, given the overwhelming pressure to just take money from both sides of the transaction and skew the decision.
But if it does happen, I actually see this as a huge potential factor strengthening smaller suppliers directly competing with large platforms. If my agent can independently figure out if a given supplier is trustworthy, whether their terms and conditions are reasonable etc., I'd be much more willing to engage with them outside of a large platform.