Turing test is bad for business

Fear of artificial Intelligence is flooded with news: unemployment, inequality, discrimination, misinformation, and even the superintelligence that rules the world. Everyone thinks that the group that will benefit is business, but the data seems to be inconsistent.In all the hype, American companies have been Slow progress in adopting the most advanced artificial intelligence technology, And there is little evidence that such technologies are Productivity growth or Create job opportunies.

This disappointing performance is not only due to the relative immaturity of artificial intelligence technology. It also comes from a fundamental mismatch between business needs and the way many people in the technical field currently conceive artificial intelligence—this mismatch stems from Alan Turing’s pioneering “Imitation Game” paper and so-called graphs in 1950. Spirit test he proposed in it.

The Turing test defines machine intelligence by imagining a computer program that can successfully imitate humans in open text conversations, so that it is impossible to judge whether a person is talking to a machine or a person.

At best, this is just a way to express machine intelligence. Turing himself, as well as other technological pioneers such as Douglas Engelbart and Norbert Wiener, understand that computers are most useful for business and society when they enhance and supplement human capabilities, rather than directly competing with us. Search engines, spreadsheets, and databases are good examples of this complementary form of information technology. Although their impact on the business is huge, they are not usually called “artificial intelligence”, and the success stories they embody in recent years have been overwhelmed by the desire for more “intelligent” things. However, the definition of this desire is not clear, and it is surprising that there are few attempts to develop another vision. It increasingly means surpassing humans in tasks such as vision and language, as well as indoor games such as chess and Go. Performance. This framework dominates both public discussion and capital investment around artificial intelligence.

Economists and other social scientists emphasize that intelligence is not only produced, or even mainly produced by individuals, but most importantly produced by collectives such as companies, markets, education systems, and culture. Technology can play two key roles in supporting collective forms of intelligence. First, as emphasized by Douglas Engelbart’s pioneering research in the 1960s and the subsequent field of human-computer interaction, technology can enhance the individual’s ability to participate in collectives by providing individuals with information, insights, and interactive tools. Second, technology can create new collectives. The latter possibility offers the greatest potential for change. It provides another framework for artificial intelligence, which has a major impact on economic productivity and human welfare.

When an enterprise successfully divides labor internally and combines different skills into a team to jointly create new products and services, the enterprise will succeed in scale. The market will succeed when it brings together different players and promotes specialization to improve overall productivity and social welfare. This is exactly what Adam Smith understood two and a half centuries ago. To translate his message into the current debate, technology should focus on complementary games rather than imitating games.

We have many examples of machines that improve productivity by performing tasks that are complementary to those performed by humans. These include large-scale calculations that support everything from modern financial markets to logistics, long-distance transmission of high-fidelity images in the blink of an eye, and classification of large amounts of information to extract related items.

The new thing in the current era is that computers can now do more than just execute lines of code written by human programmers. Computers can learn from data, and now they can interact with humans, infer and intervene in real-world problems. We should not view this breakthrough as an opportunity to transform machines into a silicon version of humans, but should focus on how computers use data and machine learning to create new markets, new services, and new ways of connecting people to people with economic returns.

An early example of this economically conscious machine learning was provided by recommender systems, an innovative form of data analysis that stood out among consumer-oriented companies in the 1990s, such as Amazon (“You may also like” ) And Netflix (“Top picks for you”). The recommendation system has since become ubiquitous and has a major impact on productivity. They create value by using the collective wisdom of the group to connect individuals and products.

Emerging examples of this new paradigm include the use of machine learning to establish direct connections between. Musicians and listeners, Writers and readers, with Game creators and players. Early innovators in this field include Airbnb, Uber, YouTube and Shopify, as well as “Creator economy“Is being used because trends are gathering. A key aspect of this kind of collectives is that they are actually markets—economic values ​​are linked to the connections between participants. Research is needed on how to integrate machine learning, economics, and sociology. In order to make these markets healthy and bring sustainable income to participants.

This innovative use of machine learning can also support and strengthen democratic institutions.Taiwan Digital Ministry Drove Statistical analysis and online participation in order to expand the scale of negotiation and dialogue to make effective team decisions in the best-managed companies.

Source link

Recommended For You

About the Author: News Center