...two different things
Using AI to improve dating apps
Using AI as a substitute for friends

Dating apps suck, and it's hard to imagine how AI could make them worse, but I can also imagine that AI could make them a lot better. AI is good at finding statistically significant patterns. With sufficiently accurate and complete data, an AI could theoretically match people better than the current scheme

AI as a substitute for friends is bad, bad, awful, no good and very dangerous. People have a history of doing bad and stupid things, so I suspect this will happen. Hopefully it will be a short-lived fad

Where expensive lawyers argue over stupid names for even worse products

...about anything coming from "a senior U.S. official " in today's political world

Future AI will be trained on every detail of product use, troubleshooting and repair. Users will be able to get highly accurate answers, as if they were talking to the designer or maintenance expert.
The pessimistic reality
Tech support is seen as a cost to be minimized and companies will gladly replace awful Indian tech support workers with even worse AI

AI generated stuff may be fun to make and listen to, and it may be a fad for a short time, but it should be honestly labelled.
The article mostly references AI generated frauds, designed to extract money from streaming services.
For those with no musical skill, tech can allow them to make stuff that almost sounds like music if you aren't too critical. It can be fun to do, but don't expect to get famous or make a lot of money.
I tried for years to use sequencers and virtual instruments to produce my compositions. At best, they sounded OK, but were missing the magic that happens when real musicians play live together.
It's also interesting to observe that when the legendary musical genius Frank Zappa tried using sequencers and virtual instruments, his results also sounded robotic and unmusical. Even when they made small mistakes, his bands always sounded better

If AI makes a product or service better, it will be accepted
If AI is seen solely as a cost cutting measure or if it makes the product or service worse, it will be rejected

That said, here's my prediction
Technologically clueless CEOs will believe the hypemongers and spend a lot of money on early AI systems
Some will work well, some OK and some will fail catastrophically
Eventually, a balance will be found, but the transition will not be smooth

Trying to use the touchscreen while moving is nearly impossible
I don't like touchscreens for anything, but when parked, they are usable
When bouncing around, even on a smooth road, it takes multiple pokes to activate functions
To make it worse, every software update moves stuff around and often makes the touch target smaller and harder to see

...useful for some things, imperfect, and rapidly improving.
Future AI is unknown and unpredictable, but this doesn't stop the hypemongers, pundits and futurists.
The general public has no facts to base their opinion on, so the opinion is worthless.

...they hype vastly exceeds reality
When billions are at stake, hypemongers use over optimistic predictions as bait to attract investment
I agree with LeCun, AGI will not be achieved using current LLM tech
New approaches will be needed

... endless superhero movies.
It's also a lack of creativity with remakes, reboots, "franchises" and other endless retelling of old stories.
Theaters are dinosaurs that deserve to die. They were once necessary because of the tech of the time, but today there is nothing pleasant about the theater experience

I suspect that most of this "music" is not made by creative people who "don't need to have fine motor skills on the piano". I suspect that most is made by scammers and mercenaries who want to harvest money by putting in minimal effort

That said, there is very little art in mercenary pop music created by teams of producers, arrangers, trendmongers, choeographers, stylists and others. It's artless industrial product, carefully crafted to be almost identical to what's popular, with just enough difference to avoid copyright trouble. It's kinda like teams of experts doing their best to imitate AI

...they might either be useful or awful

They could be useful for providing subtitles to those with hearing problems or for translating language
They could be useful for providing help or recording interesting observations
Here's one example. I am working on repairing a machine and I encounter a part I don't understand. If the glasses could scan the part, look it up and display a description of its function, it would be helpful. If it was broken, the glasses could offer prices and availability of replacements

They could become awful if they pester the user with ads
They could be awful if they were used as part of a silly, pop culture fad
They could be awful if they were used by totalitarian governments for surveillance

...increasingly unpredictable
Hypemongers continue to make inflated predictions in order to get investment or attention
Real progress is being made, but it's nowhere near the hype

Run away
AI needs local compute