Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We may not like what we become if A.I. solves loneliness

newyorker.com

Post reply on HN
Print view View original

981–990 of 1001 posts

Re: We may not like what we become if A.I. solves loneliness

#981

Post by somenameforme »

jncfhnb wrote: post #980
somenameforme wrote: post #969

Earlier quoted context omitted.

The point of the economic numbers is that what's affordable from a Western perspective is not for a Japanese person. You're talking about a country where 50% of workers earn less than $25,000 per year! A $200k house in the US is generally considered very affordable. In Japan that's 8 years of salary for half the country, and easily a 20+ year mortgage. Their housing prices are being further depressed by the fact that…

shrug. Doesn’t contradict anything I said. Japan is not a developing country.

We can argue semantics, but I specifically said that "their economy is much closer to a developing country." This contrasts sharply against, what I presume is, our youth - when they were at one point set to become the largest economy in the world, and everything Japanese was state of the art + crazy expensive while quite affordable for Japanese.

Now it's rather the opposite. If you go to Japan on a Western salary (and especially after converting Western currency post ~2022), everything's dirt cheap for a foreigner, yet quite expensive for locals, which is much more akin to the economic state when visiting a developing country.

And so saying housing is cheap in Japan is kind of crass in a way. I mean yeah obviously it is, so long as you don't happen to live and work there.

Re: We may not like what we become if A.I. solves loneliness

#982

Post by 1vuio0pswjnm7 »

Works when/where archive.is is blocked:

Disable or do not use Javascript

For example,

   x=https://www.newyorker.com/magazine/2025/07/21/ai-is-about-to-solve-loneliness-thats-a-problem

   firefox about:config javascript.enabled toggle to false
   firefox $x
   
   echo url=$x|curl -K/dev/stdin > 1.htm
   firefox ./1.htm

   links $x

Re: We may not like what we become if A.I. solves loneliness

#983

Post by ninetyninenine »

ryandrake wrote: post #917
ninetyninenine wrote: post #487

Earlier quoted context omitted.

Of course not. And neither is machine learning data. I'm talking about the trendline. If you deploy that machine learning knowledge and draw a best fit line. That line has a slope that is upward.

There were only 66 years between the Wright brothers' first successful powered flight in 1903 and the Apollo 11 moon landing in 1969. The slope of the following 66 years of aviation/flight history has been distinctly less steep. If you were in 1969 and extrapolated the trendline, we would have colonized the inner solar system by now.

Right so a single example proves your point and destroys the entire concept of trendlines. We should stop the use of trendlines immediately.

Why don’t you take a more realistic look at it. Trendlines predict things. They don’t guarantee things.

You want to make a prediction based off of bullshit? Or if you want to make a prediction based off of a trendline. I pick the later. You pick the former because of one example.

Re: We may not like what we become if A.I. solves loneliness

#984

Post by ryandrake »

ninetyninenine wrote: post #983
ryandrake wrote: post #917

Earlier quoted context omitted.

There were only 66 years between the Wright brothers' first successful powered flight in 1903 and the Apollo 11 moon landing in 1969. The slope of the following 66 years of aviation/flight history has been distinctly less steep. If you were in 1969 and extrapolated the trendline, we would have colonized the inner solar system by now.

Right so a single example proves your point and destroys the entire concept of trendlines. We should stop the use of trendlines immediately. Why don’t you take a more realistic look at it. Trendlines predict things. They don’t guarantee things. You want to make a prediction based off of bullshit? Or if you want to make a prediction based off of a trendline. I pick the later. You pick the former because of one example…

There are many, many real world examples where past performance is not indicative of future results.

Re: We may not like what we become if A.I. solves loneliness

#985

Post by vannevar »

somenameforme wrote: post #970
vannevar wrote: post #905

Earlier quoted context omitted.

The analogy of education to housing "bubbles" doesn't work. Housing bubbles are economically destructive because dropping prices induce new sellers to drop their price even further, which reduces the market value for everyone. There is a reinforcement loop. No such mechanism happens in education, once you have your degree it is yours forever. There is no secondary market. If the value goes down, sure, other people wi…

Of course there's a connection. The most realistic reason people go to college is to earn more money later in life. Nobody would ever voluntarily go 6 figures into debt with the understanding that, at the end of it all, the best job they're going to be able to find is to go serve coffee. And especially in the era of the internet the concept of college being a necessity to educate oneself is rather plainly artificial,…

>Of course there's a connection.

A connection, yes, but not a feedback loop. A drop in the value of a degree does not lead to everyone panic-selling their degree.

What you're describing is simple ROI---as the return on the education investment declines, people reduce their investment in degrees. But falling tuition does not further reduce the ROI, as in the case of housing where there is a general expectation of appreciation, and therefore, speculation.

Re: We may not like what we become if A.I. solves loneliness

#986

Post by sdsd »

ane wrote: post #399
sdsd wrote: post #278

Earlier quoted context omitted.

git diff your_argument my_response - monkeyFeelSmart(); + monkeyFeelSafe(); People feel hurt and lied to after decades of diligently studying a curriculum who's foundations turned out to be completely fake. Our mental garden must be protected from pests. Some pests even imitate benign bugs like ladybugs, in order to get in. Imagine if tomorrow, it was announced that atoms and gravity don't exist, the motion of heaven…

Could you elaborate on the subject of the foundations of a curriculum (of what?) being entirely fake? That's a bold statement to make!

I'm thinking of what the replication crisis means for figures like Jean Piaget and Marie Ainsworth. Go take a psych class and they'll be among the first you learn about. None of their stuff replicates. The curriculum I took at community college was mostly fake.

Re: We may not like what we become if A.I. solves loneliness

#987

Post by nazgulnarsil »

dv_dt wrote: post #257
stdvit wrote: post #97

> Even in a world brimming with easy distractions—TikTok, Pornhub, Candy Crush, Sudoku—people still manage to meet for drinks, work out at the gym, go on dates, muddle through real life. They actually don't. Everything from dating and fitness to manufacturing and politics is in decline in activities, and more so in effect and understanding. You can't convince (enough) people anymore that it is even important as many…

Though it's popularized to blame social media and phones, economics should not be overlooked. Pay for young generations is lagging and restaurants and bar prices are super high. Public spaces for informal gatherings has shrunk - eg fewer malls

We have eliminated the ability to discriminate against those displaying anti-social behaviors or commons damaging externalities in the name of fairness, leaving monetary discrimination the only remaining legally viable recourse and are destroying every other form of commons.

Re: We may not like what we become if A.I. solves loneliness

#988

Post by immibis »

Gud wrote: post #479
dfxm12 wrote: post #395

Earlier quoted context omitted.

This is more a function of dense population centers. Having lived in many places, I went out more in the denser areas. There are more options and they are all up and down the price spectrum. In sparse areas, going to the same few options over and over again isn't fun, and they tend to be more expensive, maybe due to lack of competition.

Don't underestimate the lack of functioning public transport. I always considered trains, tram slow teleporters. A functional rail network allow the public to move with much less restraint. Think about it. A highly car dependent society which much of the world unfortunately still is, will make going to 3rd places much less attractive. Easier to sit at home, doom scroll and watch Netflix. Inter city trains should run…

Very importantly, with public transport, you don't have to lug this huge metal box around with you, remember where it is, and be sober enough to safely operate it.

You can just go where you like, and if you want to go somewhere else, sure it might not be the strictly fastest option, but it sure is convenient. You can go from A to B to C to D to A without having to go back to B to grab your elephant box and bring it to D.

Re: We may not like what we become if A.I. solves loneliness

#989

Post by cherrycherry98 »

dv_dt wrote: post #469
andrepd wrote: post #411

Earlier quoted context omitted.

Parks and libraries are always full where I live, a metro area of 1.5mil.

The number of US libraries going back to the 90s is basically flat while the population has kept growing over 35 years, around 38% for the same time period

Unless you're a big city you probably only have one library. At least by me municipalities have occasionally built new, bigger libraries and/or put additions onto existing ones in the last 30 years. The number of libraries would remain flat in those cases even if they have actually expanded.

Re: We may not like what we become if A.I. solves loneliness

#990

Post by ninetyninenine »

ryandrake wrote: post #984
ninetyninenine wrote: post #983

Earlier quoted context omitted.

Right so a single example proves your point and destroys the entire concept of trendlines. We should stop the use of trendlines immediately. Why don’t you take a more realistic look at it. Trendlines predict things. They don’t guarantee things. You want to make a prediction based off of bullshit? Or if you want to make a prediction based off of a trendline. I pick the later. You pick the former because of one example…

There are many, many real world examples where past performance is not indicative of future results.

Of course there are. Because nothing goes up the line forever. Eventually any real world line actually ends. Nothing in the real world actually obeys the equation of a line to infinite. This happens to every fucking line in existence. So why do trend-lines even exist? Why don’t we flex some of those common sense muscles.

If we know that locally we are on a line, and we know that the trend obeyed that of a line or upward trend -x units into the past. The most probable bet is to say that the line will obey the trend +x units into the future.

The more data and the less noise we have on the past trend makes predicted future trend more likely. This is The data driven logic, no way around this other then using your gut to say random contrary shit.

My claim is given all the information we have the best most probable bet is AI will follow the trend it followed in the past. And your bet is that it won’t and you didn’t even supply any qualitative evidence for why not. You just cited common trope quotes from investing without nuance. You realize that value investing on the index sp500 is a trendline based bet?

Post reply on HN