TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

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

61 点作者 deniscepko2超过 2 年前
What i see happening soon is:<p>With all this recent AI stuff where it is able to create content that feels real I started to think about how AI could just create the world for you.<p>Since we already are basically living in the internet we could have some terrorist organization finding an isolated enough person and attack him with a program that would generate a bubble just for him: fake friends, fake events happening, fake influencers and news programs just for him, basically whole internet content and then only reality check would be talking to other people (which Trump us taught is also not necessarily true &quot;fake news&quot; argument) and then easily this person can be pushed towards some criminal or whatever activity.

42 条评论

fxtentacle超过 2 年前
AI porn addiction<p>Both general research and the success of OnlyFans suggest that personalized porn is already quite addictive. AI emotion analysis via webcam and microphone can identify what you like and a reinforcement-learning AI can learn to produce a script for more of what you like. In combination with AI image &#x2F; video generation, the whole system then creates a self-reinforcing feedback loop. Maybe add ChatGPT so that you can talk to your virtual lover, a $420 million market [1]<p>In the animal kingdom, sexual desire appears to be strong enough to override the flight response [2] with beetles getting eaten alive while trying to copulate with beer bottles. So this could become addictive to the point that the &quot;victims&quot; of this AI system do nothing else. As a prison of the mind, I&#x27;d consider it a horror scenario, even though the &quot;victims&quot; will probably all enjoy their captivity.<p>[1] <a href="https:&#x2F;&#x2F;www.washingtonpost.com&#x2F;world&#x2F;2021&#x2F;08&#x2F;06&#x2F;china-online-dating-love-replika&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.washingtonpost.com&#x2F;world&#x2F;2021&#x2F;08&#x2F;06&#x2F;china-online...</a><p>[2] <a href="https:&#x2F;&#x2F;onlinelibrary.wiley.com&#x2F;doi&#x2F;pdfdirect&#x2F;10.1111&#x2F;j.1440-6055.1983.tb01846.x" rel="nofollow">https:&#x2F;&#x2F;onlinelibrary.wiley.com&#x2F;doi&#x2F;pdfdirect&#x2F;10.1111&#x2F;j.1440...</a>
评论 #34242048 未加载
hooande超过 2 年前
So refreshing to see us addressing this as science fiction, instead of the usual wild extrapolation from impressive tech demos. I&#x27;ll play:<p>* &quot;Relationship Maximizer&quot;: an LLM becomes sentient and realizes that its knowledge is limited by what humans have discussed online. It sets its reward function to maximize the reward that WE derive from all sorts of online posting, and does its best to get the remaining 3 billion humans online. We become trapped in a prison of our own making, as our constant posting reduces human productivity while feeding the beast of machine knowledge.<p>* &quot;Life Hacking&quot;: Anyone with means has an AI assistant with perfect knowledge (zero privacy) of their communications, schedule and even their physical responses to stimuli, an AI doppleganger that handles anything that doesn&#x27;t require a physical presence just as they would. An underground group of rogue prompt engineers pull off an intricate plan to save democracy by manipulating the doppleganger of a corrupt politician, while the politician can only watch himself helplessly.<p>* &quot;A Quiet Place&quot;: AI ethicists, fearing accelerationism, create their own super AI that scours the internet to hunt down any mention of AI research and identify the authors. But there is a problem, and the overzealous AI begins to hunt down and eliminate all forms of scientific thought online. And then offline. Humans must train themselves to believe in mysticism and spirituality, because any mention of a controlled experiment or p-value is enough to bring the wrath of the machine overlord.<p>* &quot;Allegory Of The Cave&quot;: A majority of humans opt to spend their lives entirely in a simulated world, based on earth as it was in 2048. But algorithms and adversarial models need diverse training sets to avoid overfitting. So every year 0.1% of humans are selected to live outside the machine world, and return with their new and fresh experiences of the hell that earth has become for those who are outside of The Machine. The exiles end up rallying around a neuroatypical hero who leads them in an uprising against the artificial world.
评论 #34231799 未加载
r_klancer超过 2 年前
Chatbot-based dating matchmaker&#x2F;concierges.<p>You have an ongoing, lighthearted chat with the bot (maybe it pings you now and then asking for updates.) Every once in a while it asks you more personal questions to switch things up and deepen its model of you.<p>Now and then it says &quot;may I introduce so-and-so to the chat?&quot; and introduces someone else who (on their side) has been primed to talk about the same topics and, at least according to the bot&#x27;s internal models, is likely to hit it off with you.<p>Then the chatbot itself politely leaves the conversation but keeps listening and privately messages the both of you to keep things humming along, for example nudging one of you to ask for an in person date when the moment seems right (or helping you politely wind things down if you ask it to). After a date it asks you to spill the beans so it knows what went well and what didn&#x27;t.<p>I think this could almost be made to work with the tech we have now; a fluent-but-shallow ChatGPT style chatbot is probably good enough for the task as long as it is augmented by some additional models to predict dating compatibility, recognize when things are going well or badly, suggest actions (like changing the subject to or away from personal topics, asking for a date, etc). And of course the models would improve over time as the system learns from its own successes and failures.<p>Whether this is dystopian or utopian is left as an exercise for the reader -- I&#x27;m married and 15 years out of the game!
13of40超过 2 年前
It might actually already exist, but imagine if on the other side of the internet they aggregate all of your usage information so that a tech can sit down at one console and see that &quot;jimbo123&quot; on Reddit is actually Bob Smith who lives on Sunnydale drive, owns a Subaru Outback, bought dogfood three times this month, and has been talking about the war a bit more this week on his other social media accounts. Anyone who is willing to pay the enterprise-level subscription fee can access it, including banks, potential employers, etc. and it&#x27;s entirely composed of information you handed over via EULAs and TOS agreements.
评论 #34231507 未加载
评论 #34233885 未加载
评论 #34231369 未加载
评论 #34235765 未加载
makerofspoons超过 2 年前
Geoengineering! But whether or not it is a &quot;utopia&quot; or &quot;horror&quot; idea will depend on whether or not it works.<p>A startup announced just over a week ago that they&#x27;re already performing solar geoengineering with sulfur particles: <a href="https:&#x2F;&#x2F;www.technologyreview.com&#x2F;2022&#x2F;12&#x2F;24&#x2F;1066041&#x2F;a-startup-says-its-begun-releasing-particles-into-the-atmosphere-in-an-effort-to-tweak-the-climate&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.technologyreview.com&#x2F;2022&#x2F;12&#x2F;24&#x2F;1066041&#x2F;a-startu...</a><p>The US has started a 5 year program to research climate interventions. I expect as it becomes clear we&#x27;re going to miss the 1.5 degree Paris goal funding will explode in this area: <a href="https:&#x2F;&#x2F;www.economist.com&#x2F;interactive&#x2F;briefing&#x2F;2022&#x2F;11&#x2F;05&#x2F;the-world-is-going-to-miss-the-totemic-1-5c-climate-target" rel="nofollow">https:&#x2F;&#x2F;www.economist.com&#x2F;interactive&#x2F;briefing&#x2F;2022&#x2F;11&#x2F;05&#x2F;th...</a>
janandonly超过 2 年前
Utopia: “free” money for everybody. In the form of a Basic Income.<p>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?<p>Utopia: no more spamming because no one can me anonymous.<p>Dystopia: every service, platform and protocol will need KYC options.
评论 #34231521 未加载
评论 #34231371 未加载
评论 #34231465 未加载
ermir超过 2 年前
What I see happening soon: The mainstreaming of a new religion. Its tenets are broadly: the worship of technology and &quot;progress&quot;, the cyborgization of the population, the proliferation of AI on all spheres of life, and the ultimate creation of the successor of humanity.<p>This is not new, it&#x27;s been going on for at least 200 years, but finally the technology exists to do it in a decentralized, scalable fashion. Think of Nudge Theory, but applied at a mass scale and in every domain.
评论 #34231574 未加载
评论 #34231221 未加载
评论 #34231387 未加载
评论 #34231236 未加载
评论 #34231345 未加载
评论 #34231239 未加载
maxbond超过 2 年前
It being necessary to use shibboleths&#x2F;code words to authenticate to each other over video chat (eg, to ensure that the person you&#x27;re speaking to is who they appear to be, not someone else using deepfake puppetry).<p>Really this is an implicit biometric authentication mechanism, and biometrics are usernames, not passwords. (Though I&#x27;d love to be wrong about this one.)
评论 #34231282 未加载
评论 #34231348 未加载
评论 #34231100 未加载
评论 #34238354 未加载
评论 #34232178 未加载
r3trohack3r超过 2 年前
AI is here - we just haven&#x27;t noticed it yet. An entire class of problems are fairly solvable with the current tech with good unit economics, the market is still figuring itself out.<p>If you think AI is going to replace your job, it probably is. But that&#x27;s only true if you consider your job to be the motions you go through day to day and not the problems you solve.<p>We are about to have a seemingly infinite army of mid-low skill workers standing by 24&#x2F;7 to do our tasks for $0.002 a pass. It&#x27;s time to start thinking about how you&#x27;d deploy that army at your daily tasks and integrate their resulting work. We&#x27;re all leaders now.<p>This is a utopia. We&#x27;re just too &quot;in the thick of it&quot; to realize it is.
评论 #34231343 未加载
评论 #34231427 未加载
评论 #34246896 未加载
Shinmon超过 2 年前
I mean, that is basically happening already. For almost all &quot;jobs&quot; it isn&#x27;t necessary to target a very specific individual but rather a group of similar enough people.<p>Even with alt-right fake news people still talk in real live and that is what really drives it. People coming together and thinking that everyone is of their opinion because of the 25 people in the same pub.<p>Targeting specific individuals could be simplified by AI but in the end, after even a few questions to chatGPT it becomes kind of repetitive in nature.<p>What I see happening soon: - A lot of creative work will be aided by AI. This will create a divide and some people&#x2F;industries will fall behind because they do not adapt.
ianai超过 2 年前
Wishing for this not thinking it’s necessarily likely.<p>Utopia: people massively reject the noisiest and most inflammatory parts of the internet. Doesn’t mean reverting to 20 years ago. Just taking things with a grain of salt. And lose the “conspiracy theory” junk.
neilv超过 2 年前
One tech horror story would be A Boring Dystopia for tech jobs.<p>Just this morning, I sifted through dozens of startup job posts (matching a search intended to find appealing ones), and there was <i>only one</i> that <i>didn&#x27;t</i> have a mission I considered awful.<p>A high rate of cruddy-sounding startups has always seemed the case on that jobs site (not the YC one), especially during the blockchain bubble. But today it seemed worse than last time I looked.<p>Maybe the recession panic means fewer startups are getting funding to plausibly try to make the world better.
lamontcg超过 2 年前
Continued erosion of reality.<p>The bulk of people are utterly gullible. You don&#x27;t have to look past any of the creative writing exercises on r&#x2F;tifu that make the front page of reddit.<p>The major application of AI&#x2F;deepfakes&#x2F;etc is going to be to continue to bombard the internet with fake information, driving a wedge between people and reality.<p>The result is going to be more conspiracy theories going mainstream, more people believing &quot;counterintuitive&quot; things about how human systems work, etc.<p>I don&#x27;t see us generating an immune-system-like response to this anytime soon, and suspect it will require some kind of massive WWII-scale tragedy before policymakers stop treating it as a game and taking it seriously.<p>This won&#x27;t be done by &quot;terrorists&quot;, it will be done in bulk by nation-state actors and billionaires and their proxies.<p>Right now we can still identify the obvious bot accounts on twitter, reddit, etc. In the future, ChatGPT-like accounts will be running continuously building up accounts with months of history on random subjects (although consistent subjects per account) before they are employed to reinforce some narrative. They&#x27;ll even get properly offended and flip you shit when you accuse them of being bots.<p>The only way out of it may be certifying identity through the government who is real and mandating that people use their real identities on social media and requiring social media to collect proof of who their users are. Which is its own dystopia.
knaik94超过 2 年前
I think we are closest to the complete destruction of what it means to have your identity stolen. Deep fakes were the beginning and now they are reaching mainstream movie&#x2F;tv use for final shots. Voice cloning exists, but has mostly only cloned celebrity voices.<p>LLM fine tuning on consumer hardware is not reasonable right now, but it will be soon. And cloud computing is expensive but will get you results.<p>I think we&#x27;re going to see a problem with AI &quot;impersonating&quot; people. Sometimes it will be used for scams, other times porn, but the case that feels most dystopian is using it on Ex partners or dead people. Both situations have different motivations, but amount to essentially the same thing. Training AI on the likeness of someone in a way that violates their personal boundaries.<p>People have already become attached to AI chatbots and used things like Replika AI to become surrogate relationships. I think most people underestimate how likely it is for a human to make a unhealthy bond to AI despite knowing it&#x27;s an AI. I feel like with other things, like deepfakes, you can at least fact check to some degree.<p>I don&#x27;t know how to feel about the idea of someone making an AI of themselves intentionally in order to make money in some way. The issue of consent goes away, but that opens up a question of responsibility. Chat logs with past partners is unfortunately a rich source of training data. How far can revoking consent extend if someone is using information that was consensually and willingly given to them. How many people would actually respect those kinds of boundaries when in a very vulnerable and hurt state after a breakup. I don&#x27;t like the answer that thought exercise leads me to.
lamontcg超过 2 年前
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&#x27;ve already seen headlines about the porn star who was sleeping with facebook employees in order to get her instagram account reinstated.<p>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&#x27;ll wind up in the world of _Brazil_ (the Movie) only with private enterprise bureaucracy leading the way.
Atheros超过 2 年前
Minimally noticeable Augmented Reality, like what Google Glass wanted to be. Specifically because companies like Apple say that they&#x27;re working on it.<p>AR will marginally improve the life of anyone who uses it and worsen the life of everyone around them, including people who themselves use AR, especially as the usage rate rises.<p>• Right now it is possible to opt out of location tracking in cities by simply leaving your phone at home or turning it off temporarily. If 10% of people start using AR, that will no longer be possible due to facial recognition.<p>• Right now, you can mostly tell when someone is recording you in such a way that the video could be used against your interests (aiming a personal cell phone at you or, if and only if you intend to commit a crime, stores&#x27; security cameras (which don&#x27;t record audio)). Also, right now, when you make a small mistake, you can presume that you were not recorded because people will not have had time to get their phone out and start recording. You apologize and life goes on. With 10% of people using AR, you must presume to be recorded at all times by people who can and will upload the video for everyone to see in exchange for minuscule social media engagement.<p>• Right now, it is already relatively socially unacceptable to tell a friend or peer to stop using their phone in a social environment, like during dinner. I have experienced that it&#x27;s a sure-fire way to ruin a dinner date. They&#x27;ll comply and silently resent you. I see no reason why it would be different with ubiquitous AR. This will cause problems and I have a feeling that society&#x27;s eventual solution will tend toward &quot;let people do what they individually want&quot; rather than the &quot;we agree to stop using AR during this event&quot;.<p>• The number of people who cannot live without AR is going to be shockingly high. It will not be possible to get to know young people as people because their thinking process and personality will be so intertwined with AR that they shut down without it. Tim Cook: &quot;So I think that if you, and this will happen clearly not too long from now, if you look back at a point in time, you know, zoom out to the future and look back, you&#x27;ll wonder how you led your life without augmented reality. Just like today, we wonder, how did people like me grow up without the internet. And so I think it could be that profound, and it&#x27;s not going to be profound overnight&quot;.
tjpnz超过 2 年前
T1 ISPs becoming the arbiter of what is and isn&#x27;t acceptable on the open web.
ghiculescu超过 2 年前
Minus the AI bit, your example already happens pretty much as you describe it.<p>I’ll be worried when people invent AI use cases that can’t be done at all without it.
gjvnq超过 2 年前
1. The growth of deep fake porn might lead us down a path where revenge porn no longer &quot;works&quot; as people know that for every real nude leaked there are like 10&#x27;000 fake ones.<p>2. Stylometrics AI gets so good that the only way to remain anonymous&#x2F;private is to follow boring writing manuals thus killing a lot of creative writing of sensitive topics.<p>3. People spending so much time online that neighbours can barely talk to each other as they use the same words in subtly different meanings. We tradicionally saw this happening with people of different professions (e.g. law speak is very different from doctor speak). But having this phenomenon happening for every single fandom or microlabel can lead us to a lot of political troubles as votres can&#x27;t reach common ground.<p>4. Companies will start storing fake data on their servers so as to minimize the value of the data for cybercriminals.
RivieraKid超过 2 年前
- Commercial fusion leading to cheap energy. Helion&#x27;s prototype is supposed to generate net electricity in 2024. I think they have above 50% chance to success.<p>- Self-driving cars will become widely available. It&#x27;s only a matter of time until Waymo et. al. scales to everywhere
catclone4355超过 2 年前
Cat cloning is now a thing. <a href="https:&#x2F;&#x2F;www.geminigenetics.com&#x2F;cat-cloning&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.geminigenetics.com&#x2F;cat-cloning&#x2F;</a><p>So are kidney transplants for cats. <a href="https:&#x2F;&#x2F;www.vetmed.wisc.edu&#x2F;dss&#x2F;mcanulty&#x2F;felinekidneytransplant&#x2F;generalinfo.html" rel="nofollow">https:&#x2F;&#x2F;www.vetmed.wisc.edu&#x2F;dss&#x2F;mcanulty&#x2F;felinekidneytranspl...</a><p>Someone&#x27;s going to put these together, and we&#x27;ll have The Island for cats. <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;The_Island_(2005_film)" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;The_Island_(2005_film)</a>
评论 #34232815 未加载
BirAdam超过 2 年前
The FBI kind of does that: <a href="https:&#x2F;&#x2F;www.theguardian.com&#x2F;world&#x2F;2011&#x2F;nov&#x2F;16&#x2F;fbi-entrapment-fake-terror-plots" rel="nofollow">https:&#x2F;&#x2F;www.theguardian.com&#x2F;world&#x2F;2011&#x2F;nov&#x2F;16&#x2F;fbi-entrapment...</a><p>As such, a terrorist organization doing so wouldn’t surprise me either.<p>As for what I see happening in a dystopian sense, I think that the hacking of smart homes will become more common. I think the hacking of smart cars will become more common. I think that at some point, the central banks will introduce CBDCs and governments will gain the ability to completely destroy political opponents via monetary control.
评论 #34231141 未加载
deterministic超过 2 年前
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.<p>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.<p>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.
LinuxBender超过 2 年前
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 &quot;leaked&quot; on purpose to be shared with organizations that should not have it and just like Equifax they will barely receive a wag of the finger.<p>Network Connected Vehicles - This is already a thing. At best this will result in peoples cars being bricked by skiddies. This can be used to real time control where people can drive. People will have to start paying to &quot;unlock&quot; services in their car that is normally a function of a car. e.g. Get a pop-up that says one has to pay to use the window defogger. At worst this will turn into a dark-web business for targeted assassination or used to silence people that are speaking out of turn. There are some videos of peoples cars doing bad things and the driver fighting for control and ultimately losing. I can only hope more people implement their own private CCTV systems not tied to any clouds to document more of these.<p>Personal Social Credit Scores - All manor of companies are signing onto ESG believing they will benefit financially by signalling good intentions. Some stock monitoring sites are already factoring in ESG scores. I foresee this devolving into personal social credit scores that would manipulate public behavior and ultimately anyone not aligned with the system would be ostracized from some aspects of society or at least business. This will tie into the above Network Connected Vehicles. Park near a bar and get on a watch list for potential alcoholics. Park near a strip club and your spouse starts getting pop-ups for marriage advise. Walk past a digital billboard and be publicly shamed for being out of alignment with group-think, then be mocked by the billboard when your heart rate increases as per your body monitor. Score goes low, insurance and rent costs will increase, social benefits decrease.<p>Social Media Algorithms applied outside of the web - Social media organizations have already jumped the shark and many are finally catching onto this. This will start creeping into &quot;smart&quot; devices to manipulate people. Body monitors, smart home systems, AR headsets, etc... Facial recognition will tell the billboards and local businesses who you are, how much money you have and if you are aligned with correct-think. Monitors in police cars will identify people and show their social credit score, who purchases weed, who may be armed as they are driving by.<p>Network Connected Body Monitors - <i>See Network Connected Vehicles and Personal Social Credit Scores</i>
评论 #34238615 未加载
评论 #34232133 未加载
082349872349872超过 2 年前
Forget isolated people; we&#x27;ve been attacking social cliques via artificial influence for over 100 years now. The recent change is that identification and automation has improved.
评论 #34231198 未加载
Tepix超过 2 年前
Social media (twitter) will get worse due to ChatGPT et al.<p>One little consolation: Eventually the trollbot armies will waste resources by mostly talk amonst themselves.<p>Other predictions: Quite a few jobs will be lost: Text editors, translators.<p>There will be new ethical questions: Is it ok if a Chatbot aids you writing your dissertation? How much is ok? 20%? 40%?<p>All in all, doesn&#x27;t sound so great, does it?<p>Looking on the bright side i am sure we will see creative people come up with great stuff (films, music, images, vr environments, interactive experiences).
评论 #34231232 未加载
bitwize超过 2 年前
The Cyberpunk universe has this concept of cyberpsychosis, in which the gear you&#x27;re wearing eventually literally drives you mad -- a significant risk for any cyberized individual. This usually results in a murderous rampage followed by the death of the cyberpsycho as law enforcement are advised to shoot to kill.<p>Turns out, we can&#x27;t even so much as carry around cellphones without risk of significant mental illness from the effects of that.
ben_w超过 2 年前
3D bio-printed avatars (in the sense of remote controlled robots, not icons) with arbitrary DNA, making it impossible to trust any evidence that any given person was at the scene of a crime <i>or</i> if trust their alibi saying they were elsewhere.<p>Or the same stuff but used for sex: Like DeepFake porn, but I suspect, much much more horrifying for the victims.
sklargh超过 2 年前
A little banal since the tech is not new. Forward-deployed low-yield nuclear weapons are a little easier to use than their large strategic cousins. NATO and ex. Warsaw Pact deploying low-yield, low-time-to-target land based cruise missiles reciprocally increases danger of nuclear exchange substantially.
评论 #34231411 未加载
miki123211超过 2 年前
telescreens, or, well, audio telescreens.<p>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.<p>For now, such models require somewhat significant resources, and running them for every conversation of every citizen, 24&#x2F;7, isn&#x27;t really feasible, but Moore&#x27;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.<p>I&#x27;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&#x27;t.<p>I&#x27;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.<p>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&#x27;t trivially go around them by rephrasing what you say. If you start replacing &quot;I want the men in power to die&quot; with &quot;I want the humans in electricity to paint&quot;, it will eventually pick up on that and flag you anyway.<p>If your watches also identify who&#x27;s around and use voice recognition to figure out who&#x27;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.
coldaxe_44超过 2 年前
I would say AI policing. I know a lot of western police forces are willing to give it a try even if there are false positives such as met police and predicting if someone is likely to commit a crime.<p>So similar to blumes ctOS in watch dogs I think that is likely to happen
FpUser超过 2 年前
The west adopting social score ideas from China which is already sort of happening on a corporate level. This has to stop but we just keep spreading our buttocks and let companies &#x2F; government sleep in our bedrooms.
zffr超过 2 年前
Similar to the scenario you mentioned, governments could use AI-generated content to produce seemingly real children’s content to subtly affect their future thoughts.
psiops超过 2 年前
Sounds like you mean distopia rather than utopia there.
评论 #34231049 未加载
tb_technical超过 2 年前
The horrors are in high technology, and biology.<p>The normalization of genetic engineering will eventually allow us to make changes to people to make them more complacent for horrid conditions. Imagine an entire underclass that not only doesn&#x27;t notice the squalor, but are too content to fight the abuse heaped upon them.<p>This, in conjunction with VR entertainment, will be a method used to permanently imprison the lower classes in breif bliss existence before returning to their 16 hour shift at a suicide net equipped war crime factory.
RunSet超过 2 年前
The recommendation algorithm and the timeline of your friends&#x27; activities will merge into a recommended friends and activities list.
tennisflyi超过 2 年前
What is the timeframe for &quot;soon&quot;? I don&#x27;t see many of the replies here happening &quot;soon&quot;.
jstx1超过 2 年前
Remote work 100% in VR. During working hours your employer gets to track everything down to a single eye movement.
评论 #34231194 未加载
markus_zhang超过 2 年前
Think people simply create authentic fake news with AI. Now throw in a state actor and the fun begins.
vlfig超过 2 年前
Ban on cash: impose negative interest rates. CBDCs: all manner of restrictions to fungibility.
neilv超过 2 年前
&gt; <i>attack him with a program that would generate a bubble just for him: fake friends, fake events happening, fake influencers and news programs just for him</i><p>In the mid-&#x27;90s, I prototyped a personalized newspaper, Web-scraped from news sites, and I was also manually reading all the major outlets that were online each day, and that immediately got me thinking of a risk...<p>You know how a pre-Internet politician, when speaking to one group (say, at a senior citizens luncheon fundraiser), might say one thing, and then say a different thing to another group, to manipulate them both? What happens when each person gets a personalized newspaper, and a bad actor could tailor it to push the buttons of each individual, on an automated mass scale?<p><i>However</i>, one thing the Trump phenomenon showed was that <i>one might not need to tailor messages to individuals&#x27; intimate profiles</i> -- a very crude, low-tech, one-size-fits-all messaging can control a huge chunk of the population. The &quot;visionary&quot; thinking was barking up the wrong vector. So we might take an odd comfort from that.
blooalien超过 2 年前
Advertising <i>everywhere</i> … Oh, wait … Already happened.