cool hit counter 牛逼吹大了,AI被解雇了!不是说1个能顶1000个吗?_Intefrankly


No one knows if Amelia, an employee of Swedish online bank Nordnet, will be in a jobless panic. She was fired not long ago, after less than a year on the job.

Blonde, blue-eyed and professionally dressed, Amelia showed strong data analysis skills in her application, better than any known good employee, able to work with thousands of people at once, much to the expectations of the bank's CEO, Peter Dahlgren.

No one expected her to be fired. Not to mention, Amelia is an AI robot.

Actually, Amelia isn't the first AI to be fired. In the same boat as her is Fabio, a UK-based shopping guide robot. In January, Fabio was mercilessly fired from the supermarket after only a week on the job. Because it neither attracts customers, let alone assists in selling. It's said that many customers will walk around when they see Fabio. If the AI had feelings too, it would surely be cool inside.

What's the problem, and how did the AI that once clamored to replace all of humanity get smacked in the face so quickly?

First of all, yes Problems with the algorithm . While banks have huge data resources and data analysis happens to be AI's strength, without the right algorithms, they may still produce unacceptable results. Especially when dealing with highly dynamic financial operations, AI is likely to be inaccurate. Financial analysis is known for its rigor, and a difference of a hair's breadth is simply not tolerated by clients.

Secondly. Communication issues . Human language as a vehicle of communication contains a complex emotional element in addition to the logical element. If the AI's continuity of conversation and expertise is not thoroughly learned, it is likely to make jokes about communication or even answer questions that drive customers crazy. If you can't talk well, you can't sell, and naturally Fabio the guide bot will be fired.

Finally. of costs . It doesn't matter if you're AI or human, companies hire you to make money. Truly practical AI employees are pricey and far beyond the reach of most companies. Look at the huge investments made by tech giants in AI R&D and the millions of dollars a year that AI talent can earn, all of which R&D costs end up being paid for by customers. Once it became clear that the investment was not proportional to the return, it was natural for companies using AI employees to raise the axe of layoffs.

So, is AI really going to be cool? In fact, the fired AI is just an example. Another Swedish bank using Amelia, Sberbank Rossii PAO, has had considerable success, and took home last year's AI Industry Award for it. The bank currently has 12,000 back office employees and will cut to 1,000 by 2021, which they attribute to AI employees.

Why are the same AI employees headed for completely different ends? There is a variable between the fired AI and the fish-out-of-water AI, and that variable is: human intervention. It's not realistic to expect AI to take on the job completely on its own at this point. Because AI will inevitably be plagued by unique or custom problems, when it encounters questions that cannot be answered by pre-defined data, established algorithms, it must turn to human auras, and human colleagues to come to the rescue.

Of course, AI will continue to evolve and improve. There is a high level of interest in the future direction of technology in human society today, and most people believe that technology will lead the way. But in Dream Sir's view, no matter what the future holds, it should not be the technology that changes the world, but the humanity behind the technology. It is not technology that will determine the future course of human civilization, but human beings themselves, who alone are the true leaders of civilization.

Any technology is a double-edged sword, and one must be wary of its negative effects; it must be human-led and fully mobilize human capacity for reflection and anticipatory thinking in order to ensure that technology develops along a trajectory for the benefit of humanity.

Dreaming of the future

Human scale Realized value

-The first batch of incubated dual platforms go live in July inside-

1、hanlp installation and use
2、From Viacom closing its ARVR division a look at some of the top five US media conglomerates attempts at VR
4、The world from the perspective of convolutional neural networks
5、DiscuzcloseQQ Interconnection Plugin Alert Message systematically busy Please try again later

    已推荐到看一看 和朋友分享想法
    最多200字,当前共 发送