If you play more advanced, you will invite all the great gods to join the forecast army. Looking at their predictions and analyzing the key and unknown factors in these predictions is actually quite a brainstorm.

For example, I recently saw an article about 12 Alibaba's scientists in different fields, making predictions about how AI and other cutting-edge technologies will affect the world in 2018. After a brief summary, I found that these scientists have maintained a very high degree of consistency in this year's technology trends, and even some judgments are amazing.

This may be worth saying....

Overall, the 12 scientists believe that the next year, quantum computing, driverless, and blockchain will each develop. And AI technology will undergo transformation and continuous penetration in many fields during the year. The most important point is that they generally believe that as the AI ​​technology matures, the IoT field will usher in a really hot year.

Is this really the case?

Obviously, scientists' comprehensive understanding of technology is unmatched by ordinary technology enthusiasts. However, the leap between technology and commerce may be a blind spot for information on the identity of scientists. This article will attempt to summarize and analyze the 2018 technology circle events in the eyes of scientists. On the other hand, it will also analyze which factors may hinder and counteract the predictions of scientists. I hope to help readers to establish a relatively reasonable and accurate technical trend perception.

Since it is a prophetic article, then interested friends may wish to take a picture and put it a year later. If the truth is really as mentioned in this article, some new outlets are brewing out – then don’t thank me, glory belongs to scientists!

AI faces three corners: artificial intelligence transformation will become the IoT fuse?

Most of these 12 scientists are dedicated to AI and related fields, so the predictions for AI are also very rich. To sum up, their judgment on the trend of AI in 2018 can be summarized into three points:

1. The AI ​​technology of consumer commerce goes from the edge to the core, and AI enters into various sub-sectors. Machine vision is still the main means.

2. The technical focus of AI has shifted from deep learning to other areas. Technical means such as intensive learning and small data learning began to receive attention.

3. Multimodal AI connections will become mainstream. The integration of technical means such as voice interaction, machine vision, and knowledge maps in unified devices or scenarios will become mainstream.

These three judgments may mean that, in the eyes of scientists, "transformation" is about to become the AI ​​annual keyword of 2018. For example, the idea that we are used to supporting consumer products with a single technology will be broken. Deep learning, as the most important implementation of the AI ​​industry today, is also being questioned by academics and industry. Using some other machine learning techniques to make up for the lack of deep learning, awakening small costs, and the possibility of completing machine learning quickly has become one of the main themes in today's academic world.

The AI ​​with reduced computing power and data cost and mature multi-modal fusion technology will be more likely to quickly enter various sub-areas and sub-segments. On the other hand, deep transformation means that AI technology and product ideas that have become habitual in a few years may have undergone major adjustments, and many AI-based business realization methods will change. The most direct change is likely to happen in the IoT field.

After all, when the natural-style human-computer interaction maturity is extremely high and the cost is rapidly declining, IoT products will usher in a great business upside potential - imagine that IoT devices can understand and understand your various needs, completely No need for display screens and buttons, the price is very affordable, of course, it is good to sell...

Overall, the AI ​​predictions of Ali scientists are still relatively optimistic about the industry as a whole. But the question may be, is the AI ​​business temptation that has already begun, is it ready to start welcoming the AI ​​transformation in various fields? When deep learning loses its dominance and the cognitive interaction approach is fully upgraded, the AI ​​products we see today may be iterated quickly. This of course may cause industrial pain.

There is also an AI vent that deserves attention, that is, with the advancement of video understanding and AI editing technology, the video industry may become different in 2018. AI face changing, AI ad placement, AI analysis video content and other technologies are turning the video into a "big toy" that can be spliced ​​at will. The new gameplay and new markets that followed will be a big event in 2018.

The video industry may break out, and it may also mean that content creators will usher in a new enthusiasm. But the regulatory risks and moral pressures that come with it will become variables. Doing bad things with AI may be one of the key words of the year.

Scientists say that IoT will usher in the first year of the explosion.

Let us say back to the IoT that scientists are optimistic about.

The Internet of Things has been said for many years, and smart homes, wearable devices, environmental sensing devices, etc. have indeed made great progress, but replacing the Internet with the Internet of Things seems to be an unreachable thing. Its commercial market has also been delayed.

Surprisingly, scientists in the AI ​​and data-related fields seem to be optimistic about the 2018 IoT field will usher in a real explosion. This of course benefits from the maturity of AI interaction technology, which may lead to an unprecedented increase in the user experience of IoT devices. At the same time, it depends on the IO chip to continue to mature, the multiple costs of IoT will quickly reduce the efficiency of the sensor system will be greatly improved.

For the outbreak of IoT, the scientists gave a rather "stimulus" prediction. For example, Zhai Zhijie, chief scientist of voice interaction at Dharma Institute's Machine Intelligence Technology Laboratory, believes: "From 2018 onwards, the interaction between humans and machines will begin to completely get rid of any form of interaction and become closer to human interaction." Obviously, I have a strong trust in the IoT hardware supported by the AI.

To sum up the IoT predictions of scientists, they can be summarized into four sentences:

1. As the IoT hardware can understand, understand, and understand humans, its user experience will gain unprecedented speed.

2. Human-computer interaction will completely shake off the screen in 2018, and the human-machine will completely use the natural way like everyone interacts.

3. With the rise of IoT, mobile phones will no longer be heavily needed, and the mobile Internet era will be completely over in 2018.

4. As more features and better experiences come, voice-based smart IoT devices will enter our daily lives in large numbers this year.

Obviously, these pre-judgments are very radical. Probably in the cognition of scientists, the new model of human-computer interaction brought by AI has reached a very high degree of completion, and even a technical critical point can be expected. It is not difficult to realize a fully naturalized human-computer interaction by using a large number of sensing devices and integrating data such as voice, vision, motion, and temperature.

Once such a technology experience enters life, the subversive effect may be a high probability event?

Commercial lost and product silence: IoT may not be smooth sailing

But is the IoT device really going so fast?

The most important question is whether our mobile phone can really be replaced quickly? From a business perspective, this may be a relatively controversial issue.

Even if it is assumed by scientists that consumer-grade hardware that is a bit higher than current IoT hardware levels is on the market, it will be subject to potential threats from cost pressures, market perceptions, and technological advances.

On the one hand, IoT equipment has not come up with absolutely valuable consumer-grade applications. In a short period of time, there will be a big explosion out of thin air, and it will start to subvert the market structure. At present, it can only be said to be an unknown number. A very important issue is that even if the IoT device can fully perform natural human-computer interaction, what is the purpose of the interaction is still a matter to be explained. If the purpose is smart home, but the current home equipment in China is not smart enough. If it is for the purpose of entertainment, work, etc., then natural interaction does not seem to be able to replace the reason for natural interface interaction.

On the other hand, we will see that although many predictions believe that IoT will greatly invade the use of mobile phones, and achieve the effect of decentralization. However, mobile phones are paying more and more attention to integrating AI capabilities. Whether to replace mobile phones with home IoT devices, or vice versa, it is necessary to confirm the product practice. Therefore, the radical assertions of scientists at this level are likely to be resisted by different levels of external factors.

Generally speaking, although IoT devices have high saturation support of AI technology, the cost threshold is also rapidly decreasing. However, the delay in finding a suitable commercial entrance and the difficulty of surrendering a satisfactory product response will constitute two major weaknesses. This may not be the field of scientists, but it affects the efficiency of the release of industrial vents.

But in any case, IoT devices that interface and combine a large number of sensors will be very reliable in the 2018. Our well-known nouns, such as AI, IoT, quantum computing, driverless, and blockchain, are persevering in self-cultivation, a little bit of evolution and integration. This year's technology circle should at least be less lonely.