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

Ask HN: What tech utopia/horror ideas do you see materialising soon?

news.ycombinator.com

Post reply on HN
Print view View original

111–120 of 127 posts

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#111

Post by lamontcg »

All of the Kafkaesque problems that we see with account lockouts and HN being used as the support channel for problems with major tech players gets ugly and turns into outright corruption. As the social contract in the US decays it becomes more common to simply pay bribes to get employees to reinstate accounts. We've already seen headlines about the porn star who was sleeping with facebook employees in order to get her instagram account reinstated.

This will get worse as some people defend the current practices as useful economics because most of us get cheap services most of the time, for only a few people in the herd getting singled out for the Kafka treatment. There will be strong opposition in a lot of quarters from trying to impose any kind of regulations to fix these problems and they will become entrenched and just the way of living and doing business. People will start to openly talk about how to bribe tech companies as a cost of doing business as the US slowly slides into more of a third world mentality. We'll wind up in the world of _Brazil_ (the Movie) only with private enterprise bureaucracy leading the way.

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#113

Post by deniscepko2 »

maxbond wrote: post #3

It being necessary to use shibboleths/code words to authenticate to each other over video chat (eg, to ensure that the person you're speaking to is who they appear to be, not someone else using deepfake puppetry). Really this is an implicit biometric authentication mechanism, and biometrics are usernames, not passwords. (Though I'd love to be wrong about this one.)

Very interestin. I wonder if you can already have a startup based on this - providing some sort of auth. Not even sure how this kind of tech could be implemented

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#114

Post by deniscepko2 »

LinuxBender wrote: post #10

id.me - With the push to get people on id.me to pay federal taxes in the US, I can foresee this eventually being a hard requirement to do anything related to the federal government initially and then eventually required for all state government interactions. Then I foresee this data being "leaked" on purpose to be shared with organizations that should not have it and just like Equifax they will barely receive a wag o…

The personal social credit scores section i believe is already happening in China. I saw a DW documentary which was describing social scores in China.

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#115

Post by maxbond »

deniscepko2 wrote: post #113
maxbond wrote: post #3

It being necessary to use shibboleths/code words to authenticate to each other over video chat (eg, to ensure that the person you're speaking to is who they appear to be, not someone else using deepfake puppetry). Really this is an implicit biometric authentication mechanism, and biometrics are usernames, not passwords. (Though I'd love to be wrong about this one.)

Very interestin. I wonder if you can already have a startup based on this - providing some sort of auth. Not even sure how this kind of tech could be implemented

It's a social rather than technical problem. But there have been several similar startups/technologies, like web of trust and keybase.

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#117

Post by gremlinsinc »

worldsayshi wrote: post #49
janandonly wrote: post #14

Utopia: “free” money for everybody. In the form of a Basic Income. Dystopia: controls on how we spend our money. No matter if we earned it ourselves or were gifted the money. No buying or selling without it being whitelisted. Maybe in the form of CBDC’s? Utopia: no more spamming because no one can me anonymous. Dystopia: every service, platform and protocol will need KYC options.

While there's increasingly good reasons to introduce basic income there seems to be little substantial movement in that direction. I have a feeling that the reasonableness of basic income will grow at about the same pace as the power of workers will diminish. Not sure if there's a way to change that equation? Introducing a very small basic income that is planned to grow might be a good way to go about this.

I'm of the mindset and have been for awhile that the FASTER we crash, as a society --- the faster we hit rock bottom with ai taking jobs, and opportunity loss skyrocketing, the faster we get to the inflection point. Call it futuristic acceleration-ism or something akin to that, but I also feel like as bad as Trump was we needed him and his ilk to show how corrupt the govt is and can go.

He was in many ways a 'good' thing for America, by exposing the worst of America. A litmus test, or a penetration test, or something of American democracy if you will?

In the same way AI hastening up automation and job take-over will usher in UBI and post-scarcity faster than just relying on politicians to organically get there over a few decades.

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#118

Post by LinuxBender »

deniscepko2 wrote: post #114
LinuxBender wrote: post #10

id.me - With the push to get people on id.me to pay federal taxes in the US, I can foresee this eventually being a hard requirement to do anything related to the federal government initially and then eventually required for all state government interactions. Then I foresee this data being "leaked" on purpose to be shared with organizations that should not have it and just like Equifax they will barely receive a wag o…

The personal social credit scores section i believe is already happening in China. I saw a DW documentary which was describing social scores in China.

That it is. Some circles suspect that the tech originated in the US and that is a test bed to see how things play out.

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#119

Post by deterministic »

The US will continue its social decline, with the US underclass spending most of its time online, being entertained enough with fake news, conspiracy theories, games, porn etc. to not revolt.

The ruling classes will make sure that the massive underclasses will get just enough $ to stay alive and not revolt. Using minimum wage jobs, that just barely makes it possible to survive, to control the masses.

The middle class will continue shrinking and join the underclasses as AI’s and robots eventually take over most of the high paid work. Massive corporations own most of the land, controlling who gets to have a roof over their heads and who doesn’t.

Re: Ask HN: What tech utopia/horror ideas do you see materialising soon?

#120

Post by miki123211 »

telescreens, or, well, audio telescreens.

Whisper, the new Speech recognition model from Open Ai, is very, very, good with English, and really decent with other languages. With Polish for example, on a normal conversation, it gets almost everything except some very specialized words.

For now, such models require somewhat significant resources, and running them for every conversation of every citizen, 24/7, isn't really feasible, but Moore's law in AI is going strong, and that will change in a few years. Even if you want to run the largest possible model right now, an RTX 3090 can do it in faster than realtime.

I'm pretty sure that, with the amount of data the Chinese government has access to, they can build something even better than Whisper, and there are enough labor camps in China in case they don't.

I'm imagining something like a watch, or maybe AR glasses, which every citizen has to wear and which transmit everything they hear in occasional bursts, encoded with something like opus to save on bandwidth.

Once you get every conversation of every citizen transcribed and meticulously cataloged, you can do a lot of pretty interesting things. You can pass it through transformer models for sentiment analysis (to flag anti-party opinions), you can search for specific subjects (and with AI, those searches can be much more advanced than a simple keyword search) etc. Because of how smart these AI systems actually are, you can't trivially go around them by rephrasing what you say. If you start replacing "I want the men in power to die" with "I want the humans in electricity to paint", it will eventually pick up on that and flag you anyway.

If your watches also identify who's around and use voice recognition to figure out who's talking to whom, you basically have the whole social graph mapped out. One criminal slips up and gets flagged, and you can start going from there. Decrease the alarm threshold for anybody in close contact with that person, take additional signals into account, like locations which trouble often starts in, and that gives you a lot of possiblities.

Post reply on HN