Has the development of artificial intelligence gone against its "original intention"?

The wheel of history is rolling forward, and now we are entering the era of artificial intelligence. The smooth road is enough to make us optimistic and excited, but the twists and turns may also make us uneasy or worried. Facing the future, we need to adhere to three basic points: efficiency, environmental protection and well-being.

1. Accelerated integration of artificial intelligence and traditional industries

If we play a connect-the-dots game and match the six companies on the left with their recent layouts in the field of artificial intelligence, what would your answer be?

Let's take a closer look at the layout of these six companies: using AI algorithms to develop new foods; testing snack delivery robots on university campuses; launching prototypes of warehouse collaborative robots; working with Baowu Steel to promote the application of artificial intelligence in the steel industry; establishing a second artificial intelligence investment fund with a scale of US$100 million; and conducting clinical trials with British partners on the use of AI to diagnose Parkinson's disease.

The answer is this.

The reason why I share these cases is that they are very typical. These cases show that both traditional enterprises and Internet giants are promoting the integration of artificial intelligence and traditional industries, and the scope of this integration is so large that we could hardly imagine it before.

However, these cases are just the tip of the iceberg. We can see that almost all industries are accelerating their integration with artificial intelligence. So what is the degree of integration and what are the differences between industries? We can make an analysis from the perspective of venture capital.

This chart analyzes the 10 industries with the highest AI investment activity over the past five years.

Data shows that the medical industry ranks first in terms of AI investment activity. This data is somewhat expected, because it represents our current greatest expectation for AI, which is to use AI to better predict, detect, diagnose and treat human diseases, so that humans can live healthier and higher quality lives.

Of course, other industries, including traditional industries such as finance, manufacturing, energy, and logistics, are accelerating this integration.

So, from the perspective of segmented technology areas, where has the investment in the artificial intelligence industry gone?

From this table, we can see that as of the first quarter of this year, the total global investment in artificial intelligence exceeded US$80 billion, and more than half of the funds were invested in machine learning technology, including machine learning platforms and machine learning applications. This represents a major direction for the current research and development and commercialization of artificial intelligence technology.

2. Artificial Intelligence: Optimistic and exciting news

Let’s look at more application cases that may be enough to make us optimistic and excited about artificial intelligence.

The first is its application in the field of new drug development. Just in May, researchers from Harvard Medical School published a research result that they were able to increase the speed of protein folding research by one million times through AI algorithms. Although this technology is currently difficult to directly apply to new drug development, it will greatly improve the improvement of modeling technology.

The second is the application in the agricultural field. TellusLabs, an American startup, uses satellite data analysis and machine learning to reduce the uncertainty of agricultural production. In 2017, it reportedly predicted the yield of US corn crops with an accuracy of more than 99%, several months earlier than the US Department of Agriculture reached the same conclusion. This is the power of artificial intelligence.

The third is the application in the manufacturing industry. Currently, Audi can automatically and accurately identify and mark tiny cracks in parts within a few seconds through machine learning and AI neural networks. The application of this technology can be said to completely change the quality inspection process of traditional manufacturing.

Let's look at the application in the energy field. The American company HST Solar uses AI algorithms to optimize the construction plan of solar power plants, and has come to a conclusion based on its own practice: compared with the system designed by human engineers, the solar power plant designed by AI can reduce the production cost of renewable energy by 10-20%. In the solar energy industry, which requires huge investment, this is also a very big optimization.

What will be the effect when artificial intelligence is used for traffic management? Baidu is currently working with many cities to promote the construction of smart cities, including the construction of intelligent transportation systems. At last year's Wuzhen Conference, Robin Li said that the company will use AI to empower traffic management in Haidian District. With the integration of intelligent systems, traffic congestion in Haidian District will be reduced by 30%-40%. For people who have been trapped in traffic congestion for a long time, this is tantamount to good news.

Artificial intelligence has already begun to be applied in the construction industry. For example, HoloBuilder in the United States can analyze construction site materials, objects, and structures captured in 360° images to automate the construction process, especially the progress tracking. For the construction industry, which was once a "dumb, big, black and rough" industry, this kind of technology will effectively improve the efficiency of on-site management.

When artificial intelligence is applied to environmental protection, it also opens up new imagination for us. In a three-year coral reef research project in Indonesia, photographers took nearly 60,000 underwater pictures. To identify and analyze a photo, a coral reef scientist would take 10 to 15 minutes, but with the rapid analysis of artificial intelligence, it now only takes a few seconds.

After looking at so many cases, we know that the application of artificial intelligence in various industries is progressing steadily. So, from a technical perspective, what stage is the application and popularization of artificial intelligence at?

Gartner's 2018 Hype Cycle can provide some answers.

From this chart, we can see that most technologies directly related to artificial intelligence, including edge AI, deep neural network ASICs, and automated mobile robots, are still in the innovation triggering period, and only two technologies: deep learning and virtual assistants have entered the desire expansion period. Overall, the popularization of various artificial intelligence technologies is still in its early stages, and it is a phased breakthrough process.

3. Artificial intelligence: disturbing or worrying news

We have seen a lot of optimistic and exciting news in the past, but in the process of development, artificial intelligence is not a smooth road, but is destined to experience various twists and turns. Let's take a look at the news that makes people uneasy or worried.

In 2018, IBM Watson's medical department made a large number of layoffs. Speaking of which, this department of IBM has been releasing some major breakthroughs in artificial intelligence in the medical field, but this layoff has attracted global attention and doubts. Many people say that artificial intelligence in medicine is dead, but this conclusion is still premature. If we analyze it in depth, in addition to the difficulty of the technology itself, we will find that it may be that we ourselves, including the entire medical system, are not fully prepared.

Let's look at another example in the medical field. In 2018, Netflix produced a documentary called "The Truth About Advanced Medicine". This film mainly exposes that American medical device manufacturers prematurely pushed immature so-called cutting-edge medical technologies to the market, causing many medical accidents, including accidents involving surgical robots. Behind these accidents, there are not only immature technologies, but also improper operations by human doctors, or problems with human-machine collaboration, and of course the greed of manufacturers.

Let's look at autonomous driving . In the past few years, on the one hand, autonomous driving has been advancing by leaps and bounds, attracting capital injections from all sides and mobilizing all our yearning and expectations; but on the other hand, more than a dozen casualties caused by autonomous driving, even if only a dozen, have caused people's concerns. People are constantly reflecting on how safe it is when you give driving authority to artificial intelligence?

In the field of collaborative robots, we can also see the successive closures of well-known collaborative robot companies Rethink Robotics and Blue Workforce.

Also, in the field of home social robots, Jibo, Kuri, and Anki, three companies that were once considered "hot commodities" in the eyes of venture capitalists and big stars among startups, have either gone bankrupt or closed down.

Whether it is collaborative robots or social robots, what reflections can these cases give us? Is it because the technology is immature? Or is it because there is a lack of application scenarios? Or is it because the business model cannot keep up? Each question is worth exploring in depth.

We also know that Amazon has created a new market for smart speakers and has become the undisputed global leader in this field. However, such a device is also causing more and more concerns about the privacy theft of its voice assistant Alexa.

We can also see the recent apology from Apple for its AI voice assistant "Siri eavesdropping on users". These events show that while artificial intelligence brings unprecedented convenience, its negative effects, or the invisible "black hand" of the "Big Brother" behind it, also urgently need global governance and regulatory follow-up.