> Thriving companies like Arc and Linear build an entire aesthetic ecosystem that invites users and advocates to be part of their version of the world.
Afaik Arc still has no revenue and no clear path to a business model, so I’m not sure I’d call it a “thriving company.” I like and use their browser but I fully expect it to die once the money runs out, because people won’t pay for a better looking browser.
I beg to differ. Software won’t be good enough until a normal person just tells it what to do and it does what the human asked. 95% of people are completely unable to use software beyond checking email or watching YouTube or playing a game. That’s not “good”. “Good” would be to tell iMovie “make this look more like The Matrix”, or telling Expedia “book a full trip to Shanghai, with non-stop flights and hotels near such and such”. You get the idea. Let me do more, free me up so I could stop wasting time on bullshit. That’d be “good”. What we have now is mostly laughable.
When we get to where you're talking about, I think it'll still be easy to find someone who's not satisfied.
Software and technology in general have been reducing and creating bullshit in varying proportions since the start. People a hundred years ago weren't worrying about the hassle of booking a flight to Shanghai because it wasn't possible.
Yes but now that it is possible I shouldn’t be spending half a day to do it. We’re on the verge of computers _finally_ becoming full blown assistants rather than just dumb tools that you have to painstakingly learn to use and then operate by hand, and people are saying we’re “done”. Just before computers are about to become dramatically more usable for people without “computer” skills.
Why are you spending half a day? I rarely spend even an eighth of that doing tasks that people here claim to take them ages? Have people forgotten how to use computers or is the UX for those really that bad?
Because I need to not just book the tickets and an arbitrary hotel. I need to get tickets with constraints - for one thing I’d like to have a bit more legroom on that 13 hour flight. For another, I need to also sit next to my wife and kid. Hotels can’t just be anywhere in Shanghai, they need to have decent reviews, be close to the things we want to see, and not be too expensive. I also need to arrange transportation to the things we need to see, and so on and so forth. A much simpler trip to Hawaii takes hours of digging through the piles of SEO enshittified garbage. It’s like this with nearly everything you want to buy or do. But I feel like we’re ratholing on trip planning too much.
Let’s do coding shall we? How about I just tell my editor, verbally or in writing, “refactor this method so that it does this and that and then update all call sites project-wide”? Or “would invariants of this program break if I do such and such”.
Or my DAW: “add a bit of dark plate reverb to the guitar, and fade out in the end”, or “isolate guitar and slow down the track in this file, and then loop this lick, I would like to learn it; oh and also tell me where I’m not playing it correctly”. Or “give me a backing blues track without the guitar that’d fit this chord progression”.
All of that without necessarily training the model exhaustively to do every combination of these things.
We could make programs _so much_ better than they currently are, it’s not even funny.
> Thriving companies like Arc and Linear build an entire aesthetic ecosystem that invites users and advocates to be part of their version of the world.
Afaik Arc still has no revenue and no clear path to a business model, so I’m not sure I’d call it a “thriving company.” I like and use their browser but I fully expect it to die once the money runs out, because people won’t pay for a better looking browser.
> people won’t pay for a better looking browser
I’d honestly consider using it if it were paid for.
I absolutely would. I'm sure there are dozens of us. Dozens!
> Software is good enough
I beg to differ. Software won’t be good enough until a normal person just tells it what to do and it does what the human asked. 95% of people are completely unable to use software beyond checking email or watching YouTube or playing a game. That’s not “good”. “Good” would be to tell iMovie “make this look more like The Matrix”, or telling Expedia “book a full trip to Shanghai, with non-stop flights and hotels near such and such”. You get the idea. Let me do more, free me up so I could stop wasting time on bullshit. That’d be “good”. What we have now is mostly laughable.
When we get to where you're talking about, I think it'll still be easy to find someone who's not satisfied.
Software and technology in general have been reducing and creating bullshit in varying proportions since the start. People a hundred years ago weren't worrying about the hassle of booking a flight to Shanghai because it wasn't possible.
Yes but now that it is possible I shouldn’t be spending half a day to do it. We’re on the verge of computers _finally_ becoming full blown assistants rather than just dumb tools that you have to painstakingly learn to use and then operate by hand, and people are saying we’re “done”. Just before computers are about to become dramatically more usable for people without “computer” skills.
Why are you spending half a day? I rarely spend even an eighth of that doing tasks that people here claim to take them ages? Have people forgotten how to use computers or is the UX for those really that bad?
Because I need to not just book the tickets and an arbitrary hotel. I need to get tickets with constraints - for one thing I’d like to have a bit more legroom on that 13 hour flight. For another, I need to also sit next to my wife and kid. Hotels can’t just be anywhere in Shanghai, they need to have decent reviews, be close to the things we want to see, and not be too expensive. I also need to arrange transportation to the things we need to see, and so on and so forth. A much simpler trip to Hawaii takes hours of digging through the piles of SEO enshittified garbage. It’s like this with nearly everything you want to buy or do. But I feel like we’re ratholing on trip planning too much.
Let’s do coding shall we? How about I just tell my editor, verbally or in writing, “refactor this method so that it does this and that and then update all call sites project-wide”? Or “would invariants of this program break if I do such and such”.
Or my DAW: “add a bit of dark plate reverb to the guitar, and fade out in the end”, or “isolate guitar and slow down the track in this file, and then loop this lick, I would like to learn it; oh and also tell me where I’m not playing it correctly”. Or “give me a backing blues track without the guitar that’d fit this chord progression”.
All of that without necessarily training the model exhaustively to do every combination of these things.
We could make programs _so much_ better than they currently are, it’s not even funny.
Users should be able to bill a ecosystem for wasted time and hours