未来社会，人工智能将会如何影响我们的生活？

1. B. A. Rehl
, Background in computer science and hardware.
Answered 3 years ago · Author has 2.7K answers and 1.1M answer views
AI will become outdated soon and replaced with AGI. This is why I work on AGI theory and not AI which has made very little progress in 60 years.

近60年来进展甚微的人工智能很快就会过时，并且被AGI取代。这就是为什么我研究AGI理论，而不研究人工智能。

1.1 Boyan Dob
Wouldn't AI still be useful for certain specific tasks? Also, AGI will have to freely agree to work with us, no?

人工智能在某些特定的领域不是仍然有用吗？并且，AGI必须得自愿与我们合作才行，不是吗？
注释：AGI(人工通用智能AGI(Artifical General Intelligence)具有对不确定性因素进行推理，使用策略解决问题，制定决策的能力，能够自主规划和学习，使用自然语言进行沟通交流的能力)

2. Sharad Suryavanshi
, Stock market enthusiastic
Futures of AI will be brightest among all the contemporary technologies, I as an individual looking forward some of the most amazing fields where AI can establish & impart tremendous results let's dig deeper –
1.Algorithmic trading- forecasting of stocks based on chart reading with the help of machine learning neural networks ,I m very optimistic particularly about the results.

在当代所有的技术中，人工智能未来将是最光明的。随着我们不断的深入挖掘下去，我期待着人工智能能够在一些最为令人惊叹的领域里创造出具有惊人影响力的研究结果。
1.算法交易— 在机器学习神经网络的帮助下，人工智能通过读取模型图来预测股票，我对其预测结果的准确性持有乐观态度。

2.Debugging- if they can do stock sextion then why not debugging one of the wonderful experience would be to see AI in this field.
It demands very high intense hard work in AI.many years to come for this to come in surface.
3 unmanned work- I will eleborat this with respect to india.there are number of people die in railway unmanned crossing every year . AI robots can play crucial role in such situation.
Moreover they can also diffuse bombs if proper training will be provided to them.

2.选股——如果人工智能能炒股，那 为什么不试试呢？在这个领域，人工智能的应用将是一个非常美妙的体验？

在人工智能方面，还需要多年的研究，才能使得人工智能得以发挥它的作用。
3遥控作业———我将举一个例子，在印度，每年都有很多人在铁路上被撞死，而人工智能机器人在这方面可以发挥出重大的作用，进而避免很多人发生铁路交通事故。此外，如果通过训练人工智能算法，它们还可以拆除炸弹。

4.Self driving cars- This is much talked about topics.Google has done some pioneer & advanced level work too in this field.
A major breakthrough in This field can save millions of life caused by car accident.u can do ur undone work sitting behind
5 Responses based on situation- we have seen preset automated response..but in AI it will uate the condition & deliver the best possible answer according to the situation.

4. 自动驾驶汽车— 一个经常被人们谈起的话题，谷歌在这方面也做了一些开创性的工作。
原创翻译：龙腾网 http://www.ltaaa.cn 转载请注明出处

自动驾驶汽车的重大突破，一方面可以挽救由于车祸而造成的数百万人，另一方面也方便人民坐在车后面办公，完成没有完成的工作。
5 基于情景反应——前提是我们已经知道了基于某种情景下预设的自动响应。而人工智能会根据情景设置评估条件，从而提供最佳情景反应。

2.1 Joseph
Do you even think how most humans would “benefit” from the total replacement of their jobs? It will be a complete human disaster, where just an elite will (maybe?) benefit from. I cannot see the ones who put their claws on the technology to share the benefits with the rest of the population.

你有没有想过？人工智能会取代大多数人的工作，这些人又如何从人工智能中获益呢？人工智能的发展，将会是人类的一场灾难，也许只有精英会从人工智能中获益，我看不到这些精英会与大多数人分享人工智能给这个社会带来的利益。

2.1.1 Sharad Suryavanshi
This question poped up in my feed that's why I answered this..Otherwise I am also against the implications of AI in the area you mentioned.

这个问题出现在我的评论里，所以我来回答一下。但是，我不认同你所提到的也许只有精英会从人工智能中受益，而非大多数人。

2.1.1.1 Joseph
Hard to control it, once it takes a “life” of its own. We should regulate it and distribute the benefits of it to the whole population. Just saying……

一旦人工智能有了自己的“生命”，就很难控制住它，我们应该约束人工智能，并将人工智能所带来的利益分配给所有人。

3. Clark Hay
, BA and MA from Oral Roberts University (1975)
Like The Internet of Everything (IoE), Artificial Intelligence has already entered almost every aspect of modern life. It is ubiquitous. Barring some catastrophe, it will simply become more and more a part of our lives as time goes on. Already there are predictive algorithms for our preferences in buying, our most likely daily activities, what kinds of people we will be more likely to associate with, what things we will likely want need view or read. Do you use Alexa or Siri? If so then you are using AI. Google’s AI algorithms attempt to match users interests with results and the kinds of advertisements that pop up. Self-driving cars are already a (flawed) reality, and so are computer driven surgeries.

就像物联网（IoE）一样，人工智能已经进入了现代生活的每个角落，几乎是无处不在的。随着时间的推移，除非发生一些灾难，否则人工智能只会越来越多地成为我们生活的一部分。人工智能已经有了预测算法，甚至已经掌握了我们的购买偏好、我们日常生活中最可能的活动轨迹、我们更愿意与什么样的人交往以及我们最可能需要查看或阅读的内容。 谷歌的智能算法推荐系统，试图将用户的兴趣与用户的搜索结果和弹出的广告类型相匹配，这就是说明。如果你在使用Alexa或者Siri, 那么你已经在使用人工智能了。自动驾驶虽然有缺陷，也已经成为现实，这也是计算机驱动的改进，这些都是人工智能的体现。

Facial recognition, voice recognition, natural language processing, and closed circuit cameras are almost everywhere these days. Algorithms for predicting criminal behaviors are being refined as well for policing purposes, which reminds me of the “Minority Report”. Big Brother already IS a reality. The NSA is absolutely so swamped with data on all of us that its AI algorithms simply can’t sort through all it fast enough to make much of a real difference in deterring crime. Nevertheless profiles are being built at blinding speeds every second of every day and there is no end in sight. And military forces around the globe are using AI every day as well, and virtually every major corporation is utilizing AI in one form or another.

如今，面部识别、语音识别、自然语言处理和闭路摄像头几乎无处不在。为了维持治安，预测犯罪行为的算法也在不断改进，这让我想起了影片《少数派报告》，其中的景象已经成为了现实。现在生活中每时每刻都在以惊人的速度产生着大量的数据，并且看不到数据的尽头，国安局的人工智能算法，压根无法以足够快的速度对所有的数据进行排序，这使得人工智能算法无法在遏制犯罪方面发挥出真正的作用。世界各地的军事力量也在每天使用人工智能，几乎每个大公司都在以这样或那样的形式使用人工智能。
原创翻译：龙腾网 http://www.ltaaa.cn 转载请注明出处

A number of companies are beginning to develop AI programs to analyze our emotions motivations moods, and psychological make up. Some are trying to understand or mimic those aspects of the human mind that shape our wants and needs and preferences and motivations and creativity - pointing the way to some future sentient autonomous self-aware and perhaps even conscious computer based entity .

许多公司开始研发人工智能程序，借助于这些程序分析人类的产生某种活动的情绪、动机和心理构成。有些人甚至试图让人工智能程序去理解或人类的思维方式，这些思维创造了我们的欲望，需求，偏好，动机和创造力。这些人工智能程序为未来研发一些有知觉的、自主的、甚至是有自我意识的机器人指明了方向和道路。

In my own mind I find this last prospect frightening. Such a sentient autonomous self-aware and conscious computer that could teach itself would most likely pass through a “juvenile” stage where it believes it is smarter than its creators - or at least bolder. And once it “matures” it will likely perceive itself to be our superior.

如此一个有知觉、自主、拥有自我意识、能够自我教育的机器人可能会经历一个“幼稚”阶段，在这个“幼稚”阶段，这个机器人会认为自己比创造它的人更聪明，或者甚至更大胆。一旦这个机器人“成熟”，它可能会认为自己是我们的领导者。我认为，让机器拥有自我意识的这个前景是很可怕的。

Would such a machine be able to hack into every other system to control things? Would it create a host of other sentient programs to do its bidding? Would it decide to oppose us as so many movies suggest? … As silly as that may sound to some, the prospect is quite frightening to others. And once the “genie” (jinn) is out of the bottle, could it be put back in or controlled, or will it deceptively hide it’s abilities until it is too late?

这样一个拥有自我意识的机器人，有没有可能侵入其他所有的系统来控制人类生产出来的机器？有没有可能会创造出很多其他拥有感知能力的机器来执行它的命令？甚至会像许多其他电影中播放出来的那样，反对人类呢？这听起来似乎很愚蠢，但对有些人来说确实是的，可对其他人来说确是相当可怕的。可以想象，一旦“精灵”从瓶子中出来，那么它会被放回去吗？或者会被控制吗？又或者会一直选择欺骗性地隐藏自己的能力？

Elon Musk has all but given up warning world leaders of possible future problems with AI, but believes that directly connecting human brains to interface with the Internet might be a good thing. If you can’t beat AI, join with it. But becoming part of the collective Borg does not sound like the kind of a future I want to be a part of. As I have already noted elsewhere, we are already drowning in information. Having more information is NOT the answer. Having more money is NOT the answer. Having extended lives is NOT the answer. Without curiosity, the will to work hard, and the desire to do our own thinking to process and uate all of that information using well honed critical thinking skills with enhanced wisdom and compassion, none of the rest matters.
Unplug, turn off, drop out.

关于人工智能未来可能出现的问题，埃隆·马斯克几乎已经不再警告世界各国了，但是他认为，如果你不能打败人工智能，那么就加入它，将人类的大脑和互联网直接连接起来，也许是件好事。但这样的未来并不是我想要的未来。正如我在之前提到的那样，而今我们已经淹没在信息当中，可是，拥有更多的信息、金钱，寿命的延长，不是最终的目的。如果人工智能时代使得未来社会缺少了好奇心、努力工作的意愿，甚至没有了自主思考解决和评估问题的意愿、没有了经过磨练的批判性思维技能、没有智慧和同情心，那么人工智能能够给我们带来就只剩下拔掉插头、关闭和退出，这一切又有什么意义呢。

3.1 Torben Mølgaard
Regarding the AI emotions thing, I read a few years ago that Japanese authorities are actually utilizing software which in real time can assert threatening or agitated voice patterns and alx someone to take a closer look. Subways and such. Creepy stuff. And as you say, it’s crawling in, more and more places.

我几年前读到过有关人工智能情感案例，在日本，实际上正在使用一款软件，该软件可以通过实时监控威胁或激动的声音模式，实现提醒某人留心附近危险的事情。正如你所说的，令人毛骨悚然的人工智能已经爬进了越来越多的地方。
原创翻译：龙腾网 http://www.ltaaa.cn 转载请注明出处

3.1.1 Clark Hay
Agreed

我认同

3.2 Joseph
Yes, not many see it, though…

是的，虽然看的不多……

3.3 Roni Patterson
In that we humans have no idea what consciousness is, nor where it comes from nor where it’s located, how can anybody expect a machine to figure that out? Plus, machines aren’t even organic. Consciousness is now referred to as the “Hard Problem” among the science community.

在科学界里意识被称为“难题”，既然我们人类都不知道。什么是意识，也不知道意识是从哪里来，又存在哪里？那么又如何指望的上机器人能够找出这个答案呢？再说了，机器甚至都不是有机生物，哪来的意识？

4. Ben Taylor
, studied Artificial Intelligence & Startups at The Internet
AI will continue to surprise us, delight us, and scare us.

Job Reduction
Jobs that were never considered to be replaceable will see full replacement with automation. Everyone talks about professional drivers with self driving cars, but I think in the next 10 years you will see pull back on many other positions. Positions like talent assessment, appraisers, dermatologists, radiologists, pathologists, lawyers, court transcxtionists, cashiers, security personnel, psychologists, manufacturing QA technicians, and professors. Anything where a human is listening, or visually reviewing repetitive tasks. Job loss estimates will climb from the commonly referenced 7% to something closer to 15%.

人工智能持续给我们人类带来惊喜，让我们高兴的同时也让我们害怕。
工作机会减少
每个人都在谈论自动驾驶汽车的职业司机，那些曾经无可替代的工作将被自动化完全取代。但我认为，在未来十年里，将会有更多的职位减少，这些职位包括人才评估、评估师、皮肤科医生、放射科医生、病理学家、律师、法庭转录员、收银员、保安人员、心理学家、制造QA技术人员和教授。估计在听或视觉上任何有人类重复工作的地方，失业将从通常所指的7%攀升至接近15%的水平。

Delight Us

使我们感到兴奋

Safety and medical will see new records on lives lost reduced and detection accuracy increases. Did you know that in the US we average 2 kids in ERs per day and 1 dead child per month for window blind cord strangulation? Terrible metrics like this will see 1–2 order of magnitude decrease with AI. In the future every home will have 100s or even 1000s of micro cameras powered by AI. The moment anything happens in your home worthy of your review (i.e your toddler is being strangled by a blind cord) your home will contact your mobile device and escalate to neighbors if needed. On the medical side we will continue to hit high marks on detection and diagnosis. Lone wolf terrorism impact and time to intervene will continue to improve. In the future our ability to respond to something like a school shooting will allow for sub-minute control (i.e. autonomous drones with non-lethal mace/taser response ). Unnecessary deaths with deadly police force (i.e. youth in hoodie being mistakenly shot at night for holding a toy gun) will be reduced 1–2 orders of magnitude thanks to autonomous super-human threat assessments via micro-drones. Police deaths will also decrease 1–2 orders of magnitude by reducing their exposure and time-to-respond.

你知道吗？在美国。平均每天有两个孩子被送进急诊室，每个月有一个孩子死于窗帘线勒死，像这样糟糕的数字将会随着AI人工智能的发展而减少1-2个数量级，这在一定程度上减少了安全隐患，提高了医疗检测的准确性，进而减少了生命的损失。在未来，每个家庭都将拥有100个甚至1000个由人工智能驱动的微型摄像头，一旦家里发生任何危险的事情。（比如你的孩子被窗帘线勒住了），你的AI就会联系你的移动设备。并在需要的时候联系邻居。减少危险事情的发生。在医疗检测和诊断方面，人工智能将继续取得高分。独狼恐怖主义的影响和干预时间将持续得到改善。未来，人工智能在应对校园枪击案等事件方面，将会实现快速控制（例如具有非致命性的自主无人机）。利用微型无人机进行的超级人类自动威胁评估，警察的致命武力造成的不必要死亡（例如穿帽衫的年轻人晚上拿着玩具枪被误杀）数量将减少1-2个数量级，人工智能的使用减少了警察暴露和反应时间进而使得警察的死亡数量减少1-2个数量级。

4.1 Gayatri Puwar
Love delight us scenarios, to strive for. Thanks for the balancing act. Job reduction aspect not so drastic as it's nothing new, happens every decade. Scare us - now that is something we need to be careful for, all who practice it to develop with ethics and big MNCs are therefore developing principles around these, as always technology in wrong hands is devastating, not just limited to AI, think burner phones in hand of terrorists, and how mobile industry (at most, not solved completely) required Global SIM and IMSI registrations for example.

爱使我们快乐，为之努力奋斗吧。感谢你的阐述。裁员方面并没有那么激烈，只是在吓唬我们，因为这不是什么新鲜事，每十年都会发生一次。想想恐怖分子手中的一次性手机，以及移动产业（最多，没有完全解决）如何要求全球SIM和IMSI注册的事情，现在我们需要小心的是所有围绕道德与实践而发展大型跨国公司借助人工智能为所欲为，而不仅仅是限制人工智能，因为技术落入坏人手中是毁灭性的。

5. Yisong Yue
Machine Learning Professor @Caltech
Answered May 31, 2017 · Upvoted by Haris Muneer, BS Computer Science, National University of Computer and Emerging Sciences (2019)
To me, the most pressing challenge of AI in the foreseeable future is combining it with concepts from other engineering fields. As AI becomes more integrated into our daily lives, addressing issues like safety and fairness will become central to how we think about developing and deploying AI systems.
Safety Critical AI. Machine learning generally does not come with any guarantees beyond (usually fairly loose) generalization bounds. Certifying rare events is very challenging. For instance, suppose an aerospace company wants to use data-driven methods to design better aircrafts (btw, this is basically EVERY aerospace company). How can we certify that the aircraft won’t crash out of the sky with 99.99999% probability? Currently, the controller on an aircraft has extremely deterministic behavior. But as you design more flexible controllers using machine learning, you necessarily allow for non-determinism to creep into the process (or at least you can’t exhaustively quantify all the behavior in the same way as before). Similar issues arise with self-driving cars, personal robots that walk your dog, etc. In general, as we integrate AI with cyberphysical systems, guaranteeing safety becomes extremely important.

我认为，在可预见的未来，随着人工智能越来越融入我们的日常生活，人工智能最紧迫的挑战是如何将其与其他工程领域的概念相结合，解决安全和公平等问题将成为我们如何开发和部署人工智能系统的核心。因为，证明罕见事件是非常具有挑战性的，通常不会认为机器学习会超出相当宽松的泛化界限的保证。例如，假设一家航空公司想要使用数据驱动的方法来设计更好的飞机(顺便说一下，这基本上是每家航空公司的目标)。目前，飞机上的控制器具有非常确定性的行为，但当你使用机器学习设计更灵活的控制器时，你必须允许非决定论渗入这个过程(或者至少你不能像以前那样彻底量化所有的行为)，我们如何以99.99999%的概率保证飞机不会坠毁?类似的问题也出现在自动驾驶汽车、帮你遛狗的个人机器人等领域。总的来说，当我们将人工智能与网络物理系统整合在一起时，保证安全变得极其重要。

Fairness in AI. Machine learning systems learn from the data that is given to them. If the data itself is biased, then the resulting system will be as well. For instance, many industries are using data-driven algorithms to make decisions such as lending, predictive policing, sentencing for crimes, just to name a few. A key issue in these settings is maintaining fairness. Just like how humans can be biased by the information they consume, so can machines. For example, if certain neighborhoods had a high volume of crime incidences recorded, then predictive policing would send more police patrols into that neighborhood, potentially leading to an even higher amount of crime incidences recorded, which is then fed back into the predictive policing system. See also: algorithmic fairness.

人工智能是公平的。机器学习系统从提供给它们的数据中学习。如果数据本身有偏差，那么结果系统也会有偏差。许多行业正在使用数据驱动的算法辅助决策，例如，利用算法辅助做出贷款、预测警务、对犯罪进行判决等决策，这只是其中的几个例子。在这些算法设置中，一个关键问题是如何保持公平？就像人类会从日常生活的事件中而产生偏见一样，机器也会。例如，如果某些社区记录了大量的犯罪事件，那么预测性警务就会向该社区派遣更多的警察巡逻，这可能会导致记录的犯罪事件数量更高，然后反馈到预测性警务系统，这就缺失了公平。参见:算法公平。

5.1 Joseph
Like we don’t have enough competition for good jobs. AI will eliminate most of them, so all humanity will live on subsistence levels and the elite will enslave us. That unless AI will eliminate humans altogether anyway.

除非人工智能将来彻底消灭人类，否则人工智能将取代大部分人的工作，所以全人类将生活在勉强糊口的水平上，精英将奴役我们。就像我们缺乏竞争优势无法获得好工作一样。

5.2 Randy Goebel
Both Yogi Berra and Neils Bohr noted their reluctance to make predictions, especially about the future ;-) Issues of reliability and “fairness” have always been an issue in complex systems, and AI doesn’t change that. I put “fairness” in quotes because it is a complex term itself … an individual or organization’s view of fairness varies widely. But the truth is that EVERY small advance in every science (including AI, ML, nano science, systems biology) entails the need for reconsideration of engineering methods, so Yisong Yue is correct … but correct generically … whether engineers are able to adapt and exploit scientific advances is always a question, and will always be, regardless of whether the advances are in AI or elsewhere.

尤吉·贝拉和尼尔斯·玻尔都表示，他们不愿做出预测，尤其是对未来的预测;-)可靠性和“公平性”问题一直是复杂系统中的一个问题，而人工智能不会改变这一点。我之所以将“公平性”打上引号，是因为个人或组织对公平的看法各不相同，公平本身就是一个复杂的术语。但事实是,在每一个科学(包括人工智能、ML、纳米科学、系统生物学)中每一个小小的进步，都需要对工程方法进行复议,所以，岳义松总的来说是正确的……然而，工程师是否能够适应和利用科学进步始终是一个问题, 不管这些进步是在人工智能领域还是在其他领域，工程师是否能够适应和利用科学进步永远都是一个问题。