The latest technology and digital news on the web

Human-centric AI news and analysis

A astute account of how AI fits into today’s economy

There’s a aberration amid a shiny new thing and a thing that works. You just need to look at the annual Consumer Electronics Show (CES) in Las Vegas to see how much of the technology we create just doesn’t cut it and gets tossed into the dustbin of accession because it doesn’t find a alive business model.

Where does bogus intelligence stand? Recent advances in apparatus acquirements have surely created a lot of action — and fear — around bogus intelligence. Game-playing bots that beat human champions. A text-generating AI that writes accessories in mere seconds. Medical imaging algorithms that detect cancer years in advance.

How much of these abstruse advances are absolutely making it to the mainstream? How much of it is baseless hype? How will AI affect jobs? How is apparatus acquirements alteration the business model of companies?

In their book Anticipation Machines: The Simple Economics of Bogus Intelligence, advisers Ajay Agrawal, Joshua Gans, and Avi Goldfarb, answer these and many other questions and paint a very astute account of the how apparatus acquirements fits into today’s economy.

provides a very attainable and high-level overview of apparatus acquirements and the power and limits of the predictions provided by AI algorithms. The book is a must-read for business leaders and executives. But it is also a very admired study for engineers and scientists who want to accept the implications of their innovations and how the technology they create integrates into the greater economy.

The book contains plenty of abundant and useful advice and examples of how apparatus acquirements is alteration how we do things. Here are some of my key takeaways.

The power of anticipation machines

There are many misunderstandings about the acceptation and aberration of bogus intelligence, apparatus learning, and other accompanying terms. There’s are also a lot of authentic discussions about AI’s advances toward human-level cerebration and compassionate and whether aberancy is within reach or not.

But the authors of break down the accepted state of AI to a very simple albeit banal concept: prediction. “The new wave of bogus intelligence does not absolutely bring us intelligence but instead a analytical basic of intelligence — prediction,” they write.

The predictive power of apparatus acquirements algorithms charcoal the core abstraction of the book and helps us accept its effect at assorted levels.

Prediction Machines book cover
Prediction Machines: The Simple Economics of Bogus Intelligence

What is prediction? Again, the authors of Anticipation Apparatus simplify: “Prediction is the action of bushing in missing information. Anticipation takes the advice you have, often called ‘data,’ and uses it to accomplish advice you don’t have.”

Even at the most avant-garde level, most apparatus acquirements algorithms are algebraic models that adumbrate outcomes: Which class does an image belong to? What will be the value of a stock in the future? What is the anticipation that a loan appellant will default? What is the likely answer to a assertive email?

As these predictions become more diminutive and precise, they can power applications that were ahead absurd or acutely difficult, such as creating astute photos of people who never existed or developing drugs for alarming diseases.

Something you’ll hear a lot in the media is that apparatus acquirements and its accepted subset, deep learning, have been around for decades. But why has the predictive power of apparatus acquirements become such a big deal today? Most experts will tell you the availability of data and more able and cheaper accretion assets have enabled advances in deep acquirements in the past years.

The authors of take these two bounds a step further. “When the price of commodity falls, we use more of it. That’s simple economics and is accident right now with AI,” they write.

As they added explain in the book, this is a trend that has been connected in the history of accretion and technology. Computers bargain the price of arithmetic. The internet bargain the cost of distribution, communication, and search. And apparatus acquirements has bargain the cost of prediction, commodity that ahead adapted all-encompassing human cerebral labor and expertise.

“Reducing commodity to pure cost terms has a way of acid through hype, although it does not help make the latest and greatest technology seem exciting,” the authors of Anticipation Machines write.

So as far as AI is anxious today, the authors write: “Computers still cannot think, so anticipation isn’t about to become cheap.” But anticipation has become very cheap, which itself is a big deal.

How apparatus acquirements will change businesses

There are two key ways cheap predictions will change the way organizations work. “At low levels, a anticipation apparatus can abate humans of predictive tasks and so save on costs,” the authors of write.

This means radiologists reviewing more x-ray slides with the help of bogus neural networks, helpdesk operators responding to more chump queries with accustomed accent processing algorithms, and account administration systems alive more calmly thanks to apparatus acquirements algorithms admiration when and how much to stock items.

“But at some point, a anticipation apparatus may become so authentic and reliable that it changes how an alignment does things,” the authors write.

Here’s an example: Amazon currently uses apparatus acquirements algorithms to make sales recommendations. For instance, when I search for , the e-commerce giant’s belvedere uses absorption apparatus acquirements algorithms to show a list of other books that I might find interesting.

amazon book recommendations apparatus learning
Amazon uses apparatus acquirements to make recommendations.

Hopefully (for Amazon), the recommendations will argue me to acquirement not one but two books. And to be clear, Amazon’s recommendations are very decent. In fact, I often search old books on Amazon to ascertain new accompanying titles.

But at some point, the predictions will become so absolute that they will cause a major shift in the company’s business model. Right now, Amazon uses a shop-then-ship model. You make a acquirement at and the aggregation does its best to bear the acquirement to your home as fast as possible.

A altered business model is ship-then-shop: Amazon uses apparatus acquirements to adumbrate what you need, and ships it to your home. If you need the items, you buy them, and if you don’t you return them at the company’s expense. This is a model that works only if the anticipation accurateness passes a assertive beginning that makes it assisting for Amazon.

Machine acquirements and the value of data


In bookish circles, most AI analysis is focused on creating algorithms that can accomplish tasks on already-established datasets such as ImageNet, CLEVR, or SQUAD.

But in real-world applications, there are many other nuances when it comes to accepting the right data for training and advancement apparatus acquirements algorithms.

The authors of have done a great job of demystifying the economics of administration data for apparatus acquirements algorithms. “Prediction machines rely on data. More and better data leads to better predictions. In bread-and-butter terms, data is a key accompaniment to prediction. It becomes more admired as anticipation becomes cheaper,” they write.

But they also accentuate that acquisition affection data is costly and time-consuming, and active an AI aggregation involves a accommodation amid the account of more data and the cost of accepting it.

Statisticians and apparatus acquirements practitioners know that data has abbreviating allotment to scale. As you train your apparatus acquirements algorithms on more data, accurateness improvements come at slower rates. The third data point provides more useful advice than the hundredth, which in turn is more useful than the thousandth.

But things are altered when you use apparatus acquirements to run a business, the authors remind us, because from the bread-and-butter point of view, what affairs is the value you get from the prediction. So if more data improves your apparatus acquirements algorithms enough to give you the edge over your competitors (think about the move from shopping-then-shipping to shipping-then-shopping), it might be worth the investment.

This is why we see tech giants such as Facebook and Google in an arms race to aggregate data that can enhance their AI algorithms.

Business leaders must also accept that per se, having a lot of data does not necessarily put you in the right position to beforehand able apparatus acquirements algorithms. Data is split into three categories: training, input, and feedback. You need all three to beforehand and beforehand an able apparatus acquirements model for your business.

For instance, having a lot of celebrated sales annal might aggregate a good training dataset for a apparatus acquirements model that predicts sales figures. But to continuously beforehand your model’s performance, you also need the means to abduction new data (input) and analyze your fresh predictions with the actual chump behavior (feedback). This demands a business action in accession to abstruse ingenuity.

“Data and anticipation machines are complements. Thus, accretion or developing an AI will be of bound value unless you have the data to feed it,” the authors write. “If that data resides with others, you need a action to get it. If the data resides with an absolute or cartel provider, then you may find yourself at risk of having that provider adapted the entire value of your AI. If the data resides with competitors, there may be no action that would make it advantageous to annex it from them. If the data resides with consumers, it can be exchanged in return for a better artefact or higher-quality service.”

The aberration amid anticipation and judgment

machine acquirements robot

Another one of the key themes discussed in is where to draw the line amid anticipation and judgment, and where to divide labor amid AI and humans.

“A anticipation is not a decision. Making a accommodation requires applying acumen to a anticipation and then acting,” the authors write.

And this, I think, is a acute takeaway. It is important for every business leader to accept the abeyant of apparatus acquirements algorithms, but also accede their shortcomings and where they need to rely on human intelligence and decision-making.

“As apparatus anticipation more replaces the predictions that humans make, the value of human anticipation will decline. But a key point is that, while anticipation is a key basic of any decision, it is not the only component,” the authors of Anticipation Apparatus write. “The other elements of a decision—judgment, data, and action—remain, for now, firmly in the realm of humans. They are complements to prediction, acceptation they access in value as anticipation becomes cheap.”

Judgment is a complicated task, often requires commonsense and compassionate of the world, two areas where apparatus acquirements algorithms currently struggle. In many cases, humans must judge and decide across assorted objectives that span across the short- and long-term. They must assess activating situations and appraise tradeoffs. But in less complicated environments, acumen and controlling can be automatic through reward action engineering or the accomplishing of hard-coded rules. Those are areas where branches of AI such as accretion acquirements might be able to fully automate tasks.

Understanding the apparatus acquirements business model

If there’s one thing that highlights, it is the axiological differences and challenges of active AI businesses. Many companies and business leaders come with a accomplishments in archetypal software development and business. They’ll need to adapt to the rules that govern the development and aliment of apparatus acquirements models, and manage the unique risks that come with it.

Those who adapt to the business of bogus intelligence are bound to reap the rewards. Those who don’t will be in for awful surprises.

This commodity was originally appear by Ben Dickson on TechTalks, a advertisement that examines trends in technology, how they affect the way we live and do business, and the problems they solve. But we also altercate the evil side of technology, the darker implications of new tech and what we need to look out for. You can read the aboriginal commodity here.

Appear September 14, 2020 — 14:00 UTC

Hottest related news

No articles found on this category.
No articles found on this category.