[soared]

People probably made that claim about a ton of other things which turned out false. The best example is music discovery, older generations might've thought "I want to pick my own music, there is no a way a computer could know what I like". But then show them Spotify's discover weekly and they discover 10 songs a week that they live.

[pasabagi]

I think Dijkstra made a point that when somebody sees something new, they try to project something they know onto it - so with computers, the first thing people tried to do was make machines that thought like people- a project that's still bearing very little fruit. In the meantime, the things you could do really well with computers, like discrete maths, have been fabulously useful.

I suspect robots are a bit like that. They make very poor substitutes for people. But, they can do a lot of stuff ('dishwashers', for example), that are basically orthogonal to what people do, but reduce workload immensely.

[ghaff]

Nah. Older generations aka me listened to the radio which is just a much cruder form of discovery. You pick a station or stations that appeal to your tastes and let the DJ do the rest.

[mynameisvlad]

There's still a difference between hand-curated content and algorithmically generated ones. People have trust in others, people are less likely to trust code.

[tjr]

A couple of weeks ago, I heard on the radio a program, something like, "The Frugal DJ", wherein all of the music played came off of records that the host had obtained for cheap. Like, $1 albums at garage sales.

It was interesting.

[misterprime]

>People have trust in others

A history of mafia men shoveling cash and threats to DJs has me convinced otherwise.

There is/was big money to be made getting your song played on the radio and that system is/was gamed pretty hard.

[gowld]

The average person has no idea whether the content on the radio is human generated (by what human? a corporate marketing department!) or algorithmically generated.

[wpietri]

People make all sorts of bad predictions in both directions, so the fact that some people were wrong on this doesn't say much about other claims. But if one is looking for a statistically safe bet, "people don't want proposed idea X" is a pretty good one. Most ideas don't work out.

[tokyodude]

Sadly I keep getting told that but have yet to find any music service that actually figures out what I like. My fiction is that you have to like one of 5 or 10 mainstream styles or you have to not be picky.

I put in Prince and Spotify gives me rap. Prince has nothing to do with wrap. I put in Pizzicato 5 and Spotify gives me J-POP. Pizzicato 5 has nothing to do with J-Pop. I put in a dance tune from a popular western pop band and Spotfiy gives me ballads from popular western pop bands when what I wanted was more dance tunes.

I've never had Spotify or Apple Music or Youtube Music work ever.

Even their own playlists don't work. I tried playing "Morning Pickup" and first 2 songs are super depressing.

[imtringued]

Last time I used Spotify's discover weekly is was completely useless. The recommendations consisted of mostly the same super specific sub genre of music with zero variety that I don't like at all.

[atomical]

So if a human was given a choice of a robot that made dinner and washed dishes and a robot that lets them check the temperature and play despacito to their hearts content they would choose the latter? I'm not buying it.

There are so many gains to be made with real robotics.

[bduerst]

I.e. Nobody would admit to wanting an camera+phone+email device before the first iPhone was released.

[ergothus]

Give me a smartphone: "This is great, I don't need a watch!"

Give me a smart watch: "This is great, I don't need to pull my phone out!"

Predicting what people will/won't like seems like black magic to me, with after-the-fact efforts to diagnose why "obviously" people did/did not love something somehow only works in hindsight.

I'm not saying the idea of finding something useful and making money providing it is broken, only that pushing the boundaries is inherently unpredictable, because we aren't terribly logical (or at least have a lot of variables at play)

[gowld]

> Give me a smartphone: "This is great, I don't need a watch!"

Who ever said that, and why didn't those people carry pocketwatches?

[jacobolus]

> Who ever said that?

When I was a kid in the 90s, my impression was that most adults wore cheap quartz wristwatches. 25 years later, their kids have grown up but instead of wearing watches just carry phones around, because the smartphone is a good enough time-keeper.

Obviously many people still wear wristwatches, and the same type of people who wore watches as status-marking jewelry 20 years ago probably still do today. But wristwatches are no longer necessary for telling time, as pretty much everyone carries a smartphone everywhere they go. For most people the additional convenience of having the watch on the wrist isn’t worth carrying an extra device, at least not all of the time.

[ergothus]

I did. Digital watches were cool! But then then were bulky blocks of plastic on my wrist, and when I was carrying a phone anyway I did away with the watch.

Years later I got the original Pebble, and while it was great for seeing who was calling me or getting a quick text, or banishing phantom buzz...my primary use was to check the time.

[ghaff]

Lots of people did/do. Smartphones added enough functionality that they became worth carrying around. It's not that they wanted to carry something around but smartphones made it worth it. And, once you have a clock with you at all times, why do you need a watch?

(To be clear, I've always liked having a watch on my wrist but then I grew up with one.)

[photojosh]

I had a Palm Treo, 3 years before the iPhone. Even had TomTom installed for GPS mapping. It was rubbish at all of those things compared to today's devices, but I could see the potential.