BAT giants are uncertain about the artificial intelligence start-up companies just pay for bubbles?
Artificial intelligence currently depends on massive data, algorithms, and computing power to drive it. But the most important thing is that the application of artificial intelligence has not been affected by the logical understanding ability and subversive business model based on the user's specific scene requirements.
However, at present, the most important aspect is that the application of artificial intelligence has no impact on the business model of logical understanding and disruptive business based on user-specific scenarios.
In Silicon Valley, Google, Amazon, and Microsoft all launched their own artificial intelligence infrastructure, APIs, and open source frameworks, including computer vision, voice, language, knowledge maps, and search. The artificial intelligence assistants of the Silicon Valley tech giant have basically become standard: from FacebookM to Amazon Echo, from GoogleAssistant, to AppleSiri, and IBM Watson.
However, even these giants' artificial intelligence assistants have their service scope basically located in information retrieval and information collection, but they can't get relatively complicated problems.
For example, these assistants can basically answer the weather today, but if they ask whether the nearby Starbucks can use WeChat to pay and whether today's weather will cause such logical problems as traffic congestion or flight delays, they cannot do anything about it.
In the level of reasoning, logic, and professional practical problems, it is still ridiculed by some users and industry professionals as intellectual disability assistants, and its practical value is not great.
Overall, judging from the strategic layout of giants, iOS and macOS require SIRI to find breakthroughs in hardware and software operations; Cortana is attached to windows; echo is associated with the operation of smart home devices. However, at present, it still does not have the same ecological and hardware entrance conditions, but it is only an ecological supplement.
In addition, the AI ​​assistants developed by many giants are lacking in the context-based dialogue ability and understanding of the logic in spoken language, ability fulfillment, and context-based dialogue capabilities, which means that the current artificial intelligence is still in a relatively elementary stage.
Artificial intelligence still has many problems. It is also difficult for entrepreneurs to fight talents, users, traffic and capital with giants.
So, this wave of artificial intelligence, which focuses on deep learning, has been blowing for many years. The giants have also invested a lot of resources. However, from the current giant's artificial intelligence assistants, it is clear that deep learning has many problems in dealing with complex tasks. Insufficient, that is to say, deep learning technology currently lacks the ability to logically reason and express causal relationships.
Obviously, if we can't make a multi-level deduction based on logic and understand the ability to express causality, we can't play a deep service. This is why we see that the chat robots developed by the giants can only make a few rounds of dialogue.
In addition, artificial intelligence currently has many technical problems that still need to be addressed. From the current point of view, in addition to conventional hardware carriers such as mobile phones and computers, artificial intelligence has not yet reached a relatively mature new hardware and software carrier. Human-machine voice interaction The degree of intelligence is low, and there is a lack of support at the hardware level. Speech recognition, natural language understanding and other technologies to commercialization of product landing is still not see signs.
Even at present, artificial intelligence has made great progress compared with the past, but its main application is still in the field of enterprise services. The artificial intelligence applications that users can access are still mobile phone and computer-based voice assistants. Although the functionality of service robots for artificial intelligence applications continues to improve, the current level of product experience is still far from commercialization and consumers.
Moreover, artificial intelligence can not be separated from the support of massive data. For entrepreneurs, once the tech giants are driving their own horsepower and increasing their firepower, when it comes to the layout of artificial intelligence, compared with the giants, startup companies are not in a grade in talent reserve and data, users, traffic, and capital. Entering this track, the result is imaginable.
At present, talents in the field of artificial intelligence are rare talents, and it is difficult for startup companies to grab quality talents.
Li Kaifu, an innovation workshop, once pointed out that companies in Silicon Valley can give doctorates in the field of artificial intelligence that they have just graduated more than 2 million to 3 million US dollars in annual salary. This kind of talent competition also leads entrepreneurs to be restricted everywhere in the process of rushing with giants.
The nature of artificial intelligence is spelling technology: but it is hard for entrepreneurs to fight giants.
Even the idea of ​​a bigger acquisition is quite dangerous, because artificial intelligence is essentially a fight for technology, and most of the current Internet entrepreneurial successes are based on business model innovation.
It is very difficult for start-up companies to make absolute technical barriers in a vertical area. Therefore, the industry has mentioned such a case. After a large Silicon Valley company acquires an artificial intelligence startup, it finds that various indicators and performance are not as good as internal ones. Products, so the team was purchased to send all the products.
In addition, after Google released a new version of the neural machine translation system, a certain positioning in the machine translation entrepreneurial team found that the accuracy of the translation of their own products lags behind Google. This shows that it is very difficult for entrepreneurs to fight technology to fight giants. On the other hand, it is even more difficult for an AI startup company to increase the probability of being acquired compared to other companies that fight business models.
Power Meter,Energy Meter, Best power meters,Cycling Power meters,Electrical power meter
NINGBO COWELL ELECTRONICS & TECHNOLOGY CO., LTD , https://www.cowellsocket.com