考虑到繁荣和萧条周期的运作方式，人工智能冬天的问题不是是否会出现，而是何时出现
Given how boom and bust cycles work, It's not a matter of if but when AI winter is coming. Once Big Tech gets bored of AI, what does /g/ predict to be the next big thing in the world of tech?
考虑到繁荣和萧条周期的运作方式，人工智能冬天即将到来不是是否到来的问题，而是何时到来的问题。一旦大型科技公司对人工智能感到厌倦，/g/ 预测科技世界的下一个重大事件是什么？
>>109590893
Youre insane if you think AI is going away at this point.
如果你认为人工智能此时就会消失，那你就疯了。
What will happen, is we reach a cross roads of local models becoming good enough, and cost of components dropping enough for people to afford dedicated AI boxes.
将会发生的事情是，我们会到达一个十字路口：本地模型变得足够好，并且组件成本下降到足以让人们买得起专用的人工智能盒子。
This wont happen anytime soon, thanks to trumpian economics. BUT hopefully in a few years, when we unfuck ourselves, its possible.
由于特朗普经济，这种情况不会很快发生。但希望几年后，当我们摆脱自己时，这是可能的。
>>109590893
>next big thing in the world of tech?
>科技界的下一件大事是什么？
What next big thing? AI is *the* information technology. Everything else is just side quests.
下一件大事是什么？人工智能是*信息技术。其他一切都只是支线任务。
>>109590893
AI is putting Math Professors out of work.
人工智能正在让数学教授失业。
There is no Winter. You are getting replaced.
没有冬天。你正在被取代。
>>109590893
3D TV, obviously
显然是 3D 电视
>>109591020
Ya it was Trump that dumped 10 trillion dollars over night and created for the first time since the 1800s 18 month treasury bonds. Oh wait no that was Biden and Janet Yelen. I hate dumb as rocks liberals who don't know about economics.
是的，特朗普一夜之间抛售了 10 万亿美元，并自 1800 年代以来首次发行了 18 个月国债。哦等等，不，那是拜登和珍妮特·耶伦。我讨厌不懂经济学的愚蠢的自由主义者。
>>109591105
He sure didn't do anything to fix it except enrich himself and his friends through insider trading, but go ahead and fellate whichever elite you want
除了通过内幕交易让自己和他的朋友们致富之外，他肯定没有做任何事情来解决这个问题，但是尽管去口交你想要的任何精英吧
>>109591121
Trump or Biden isn't the reason your broke
特朗普或拜登不是你破产的原因
>>109591020
Away? Nobody said away. But despite heroic injections of public funds into AI, they can't make any money. In fact their creditors are frowning and even NVIDIA is frowning at companies they leased nearly free GPUs to and reported said leases as sales, in a scam that should be illegal but somehow isn't.
离开？没有人说走开。但尽管公共资金大力注入人工智能，他们却赚不到钱。事实上，他们的债权人对此感到皱眉，甚至 NVIDIA 也对他们向其出租几乎免费 GPU 的公司感到皱眉，并将所说的租赁报告为销售，这是一个应该是非法的骗局，但不知怎的却并非如此。
Somebody is going to have to make some money at some point aside from NVIDIA or else the entire tech-o-sphere collapses. Semiconductor factories are booked out 'till 2030 on promises that the money is "coming soon!" and the same with the investor bag-holders. It's not sustainable, which is why these companies are all shoving AI into everything and jacking up the subscription costs and spy-monetization AI enables. A desperate search for revenue so they can pay the AI companies they've partnered with, so these AI companies can pay the datacenters, so the datacenters can pay the computer and board makers who can pay the semiconductor people. The feeder, the AI consumer market / great replacement / whatever it is? If that doesn't turn into MORE MONEY than uptake will be shit and the rest of the downstream industries get fucked.
除了 NVIDIA 之外，在某个时候必须有人赚钱，否则整个科技圈就会崩溃。半导体工厂 2030 年之前的订单已被预订一空，并承诺资金“很快就会到账！”投资者的持有者也是如此。这是不可持续的，这就是为什么这些公司都将人工智能推向一切，并提高订阅成本和人工智能的间谍货币化。他们迫切地寻求收入，这样他们就可以向与之合作的人工智能公司付款，这样这些人工智能公司就可以向数据中心付款，这样数据中心就可以向计算机和主板制造商付款，而计算机和主板制造商又可以向半导体人员付款。喂食器、AI消费市场/大替代/不管是什么？如果这不能转化为比吸收更多的钱，那么其他下游行业就会完蛋。
>>109591049
math professors' primary contribution to the economy is grant writing, teaching, and grading, they aren't being put out of work until openai opens a university
数学教授对经济的主要贡献是资助写作、教学和评分，直到 openai 开设一所大学之前，他们不会失业
>>109591198
>. But despite heroic injections of public funds into AI,
>。但尽管公共资金大力投入人工智能领域，
IDK where this idea comes from but the US government has been completely restrained in giving much money for any AI companies.
我不知道这个想法是从哪里来的，但美国政府在为任何人工智能公司提供大量资金方面一直受到完全克制。
What money has been given is completely dwarfed by private investors and companies themselves.
所给予的资金与私人投资者和公司本身相比完全相形见绌。
I guess the confusion comes from whenever the trump admin announces investments but that's not at all tax money but basically "words of intent" from outside investors.
我猜这种混乱来自于特朗普政府宣布投资时，但这根本不是税收，而是外部投资者的“意向书”。
If you twist yourself into a knot I guess it's possible to claim that tax breaks from all those involved is "public" money but you'll have to be outraged about alot more things than a datacenter.
如果你把自己陷入困境，我想你可以声称所有相关人员的税收减免都是“公共”资金，但你必须对比数据中心更多的事情感到愤怒。
>>109590893
Robotics will have a (proper) boom but that is in some years time (5-15 I reckon, but probably in the tail end). Brain chips and brain interfaces will be massive at some point, hopefully many years away, pray with whatever means you have that we will live in some utopia till then
机器人技术将迎来（适当的）繁荣，但这需要几年时间（我估计是 5-15 年，但可能是在年底）。大脑芯片和大脑接口在某个时候将会变得巨大，希望是在很多年之后，用你拥有的一切手段祈祷我们在那之前将生活在某个乌托邦中
>>109591020
Profitability will not be reached, prices will go AWOL, then with time it will become a cheap nifty tool like everything else
无法实现盈利，价格会擅离职守，然后随着时间的推移，它会像其他东西一样成为一种廉价而漂亮的工具
They're gonna put all their money into buying territory/citizenship in third/second world countries. Vacation homes, bunkers for the elite, that sort of thing since that 1%-10% hold like 90% of the wealth. Pleasing the bottom 90% doesn't matter anymore, they don't want the volatility that comes with a fragmented nation, so they'll make their little club countries invite only and use complacent Rednecks/Chinese/Indians to do all the time consuming crap.
他们会把所有的钱都用来购买第三/第二世界国家的领土/公民身份。度假屋、精英阶层的掩体等等，因为 1%-10% 的人掌握着 90% 的财富。取悦底层的 90% 已经不再重要了，他们不希望国家分裂带来的波动，所以他们会让他们的小俱乐部国家只受邀请，并利用自满的乡下人/中国人/印度人做所有耗时的废话。
>>109591220
AI can do that, lol.
AI 可以做到这一点，哈哈。
>>109591146
Lick that boot
舔那只靴子
>>109590893
>What's next?
> 接下来是什么？
HAL is to IBM what AI is to BJ
HAL 之于 IBM 就像 AI 之于 BJ
>>109591105
yes, it indeed was trump that shut america down because of the flu
是的，确实是特朗普因为流感而关闭了美国
>>109590893
It's not gonna be easy to keep number going up. I'm thinking (fake) alien tech.
保持这个数字的上升并不容易。我在想（假）外星科技。
>>109591020
It doesn't need to "go away" for a major crash or crisis to happen. The Great Depression didn't eliminate banks. The Oil Crisis of the 70's didn't eliminate oil. Dot Com bubble didn't end the Internet.
它不需要“消失”就会发生重大事故或危机。大萧条并没有消灭银行。 20世纪70年代的石油危机并没有消除石油。互联网泡沫并没有终结互联网。
AI's plateauing is increasingly obvious. The big gains are all behind us, what's left are improvements so small they're not worth the investment.
人工智能的停滞期越来越明显。巨大的收益已经过去，剩下的改进很小，不值得投资。
"AI write me a book/script" and "AI generate me enough videos for a movie" are what is being pushed right now, and eventually that will fall flat when it fails to make a profit.
“人工智能给我写一本书/剧本”和“人工智能为我生成足够的电影视频”是现在所推动的，但最终当它无法盈利时就会失败。
>>109592273
>AI's plateauing is increasingly obvious.
>人工智能的停滞期越来越明显。
Is it? It feels like the opposite. Its only just recently became useful. New math theorems are being proved every day. Vibecoding just werkz...like ridiculously well. Earlier today I asked Codex to edit my KOTOR executable directly to make it compatible with 1440p widescreen resolution and...it just did it. In one shot. In only a few minutes.
是吗？感觉恰恰相反。它最近才变得有用。新的数学定理每天都在被证明。 Vibecoding 只是 werkz... 好得可笑。今天早些时候，我要求 Codex 直接编辑我的 KOTOR 可执行文件，使其与 1440p 宽屏分辨率兼容，而且……它就这么做了。一击。只需几分钟。
>>109591337
AI cannot provide the degree with the name of a credentialed institution as the issuer, which was the whole point of
AI 无法提供具有认证机构名称作为颁发者的学位，这就是该学位的全部要点
>until openai opens university
>直到openai开办大学
>>109591020
yeah we just need to get AOC in there so she can wave her magic wand to make this all go away then we can go to brunch
是的，我们只需要让 AOC 进去，这样她就可以挥舞魔杖让这一切消失，然后我们就可以去吃早午餐了
>>109591020
You’d be insane in 1999 to think the internet was going away, but the Dotcom bubble still burst didn’t it?
如果你在 1999 年认为互联网即将消失，那你一定是疯了，但互联网泡沫仍然破裂了，不是吗？
>>109590893
Back to Bitcoin
回到比特币
>>109591049
>HURR DURR U GIT REPLACE HURR DURR
I hate this retarded FUD so much.
我非常讨厌这种迟钝的 FUD。
>>109592697
meanwhile, dead internet theory
与此同时，死亡的互联网理论
>>109590893
Non-personal computing. Everything will be run by AI remotely.
非个人计算。一切都将由人工智能远程运行。
>>109591020
>cost of components dropping enough for people to afford dedicated AI boxes.
>组件成本下降得足以让人们买得起专用的人工智能盒子。
You really don't understand do you? The people who control the world (not Trump) are making sure this will never be possible. They need to make compute power a scarce commodity like land and oil. Oil is gradually being deprioritized for various reasons, land they are going pretty hard on behind the scenes to get under corporate control, and compute power is a new concern they are just trying to get a handle on with fake shortages and artificial price inflation as a temporary stop-gap before more drastic action is taken soon.
你真的不明白吗？控制世界的人（不是特朗普）正在确保这永远不可能。他们需要使计算能力成为像土地和石油一样的稀缺商品。由于各种原因，石油逐渐被取消优先地位，他们在幕后非常努力地争取土地以使其受到公司控制，而计算能力是一个新的担忧，他们只是试图解决假短缺和人为价格上涨的问题，作为很快采取更激烈行动之前的临时权宜之计。
>>109591337
Ok, then where are the AI universities?
好吧，那么人工智能大学在哪里？
robotics
we've had years of improved hardware designs, sensors, and how they communicate
我们多年来一直在改进硬件设计、传感器以及它们的通信方式
now the software side is rapidly filling in some gaps in learning and autonomy
现在软件方面正在迅速填补学习和自主方面的一些空白
the momentum is building
势头正在增强
>>109592306
>ai is going to replace humans because i told it to do [extremely simple thing] and it did it
>人工智能将取代人类，因为我告诉它做[极其简单的事情]，它做到了
>>109592789
Won't take off in America because of certain people we allow to roam around freely.
由于我们允许某些人自由漫游，因此不会在美国起飞。
>>109592789
also, robo dogs and autonomous drone swarms are quietly joining military ranks. Military leads tech development, then it trickles down to civilians
此外，机器狗和自主无人机群正在悄然加入军队行列。军事引领技术发展，然后惠及平民
>>109592789
forgot to mention AI is what's fueling the software side improvements
忘了提人工智能是推动软件方面改进的动力
good match, honestly
很好的匹配，老实说
>>109592789
right when I posted that, Ubitree IPO mooned. screencap this for future generations to know of my powerful precognition: it has begun
就在我发布这篇文章的时候，Ubitree IPO 正在酝酿之中。截屏让后人知道我强大的预知力：它已经开始了
>https://www.cnbc.com/2026/08/19/china-backflipping-robot-maker-unitree-jumps-shanghai-ipo.html
quantum computing will open up some novel applications that generate hype, but not for another 15+ years
量子计算将开辟一些引起炒作的新颖应用，但还需要 15 年以上的时间
materials science, physics, manufacturing, and algorithmic theory each need to improve, and that is happening, but slowly
材料科学、物理学、制造和算法理论都需要改进，而且这种情况正在发生，但速度很慢
>>109591020
this is the exact same shit that has been repeated during the shitcoin and nft scams
这与垃圾币和 NFT 骗局中重复出现的一模一样
>>109597194
??
crypto and NFTs are distributed technology. Nobody was talking about running your own blockchain in a few years
加密货币和 NFT 是分布式技术。几年内没有人谈论运行自己的区块链
（如果你觉得这篇文章有启发，可以点击这里付费）
