The phenomenon of AI is becoming increasingly widespread: At every gathering or conference and in every classroom people seem to be talking about it. The news are also covered with headlines regarding AI, most of them predicting that it will completely change the ways we live and work. Yet, there is also a lot of hype over it simply for the sake of ‘buzz sells’, a danger which many scholars are warning about.
Overhyped AI can be dangerous
Countless companies and governmental organizations are raising funds on behalf of AI, which is why billions of capital run into AI start-ups. However, many of these start-ups proved to be economically unsustainable. “People who can do it have no opportunities and resources. On the other hand, people who should not do it waste resources because they are not interested in advancing technologies but simply want to grab some money,” said WANG Feiyue, who is a specially appointed Chinese state expert, at this year’s IEEE International Workshop on Artificial Intelligence and Cybernetics. The hype over AI has created an illusion that can confuse people and misrepresents what AI is really capable of. At JANZZ, we have discussed this issue in a previous article which draws on illustrating cases. Our research leads us to assume that the hype about big data and AI is often more about self-marketing instead of a focus on facts and progress. This can be dangerous, not least because it stimulates workers’ anxiety about being replaced by AI. For further elaboration on this topic, visit https://janzz.technology/even-ado-nothing-hype-big-data-ai-often-self-marketing-facts-real-progress/.
What AI is really all about
In public discussion, AGI (Artificial General Intelligence) and specialized intelligence are frequently confused and both referred to by the term AI, says Andrew Ng who is the founding lead of the Google Brain team, a former director of the Stanford Artificial Intelligence Laboratory and the overall lead of Baidu’s AI team. AGI refers to human-level intelligence; that is, the kind of futuristic intelligence that we see onscreen and in science fiction literature. Currently, the technology to reach such “human intelligence” is very limited and still far from being beneficial to society. Most of what is covered in the media refers to specialized intelligence such as machine learning, computer vison or natural language processing. Specialized intelligence is thus the real force in the fourth industrial revolution, as it creates value and bears transformative power for all industries.
Although AI has already an impact on many industries — web search, finance, and logistics, to name just a few — its subtypes that are being developed are still quite limited. Andrew Ng explains that almost all the recent progress of AI is thanks to a simple A (input data) to B (simple response) process type called “supervised learning”. Some example to illustrate this: you show pictures (A) to the software and it can identify whether it shows a cat (B); you give both ad and user information to the software (A) and it tells you whether the user is likely to click on said ad (B). Arguably the best development based on the A-to-B type is the so-called deep learning, deep neural networks that are inspired by the human brain. However, there are two crucial factors playing into the functioning of the A-to-B relationship. One is that A and B have to be carefully chosen; that is, to provide the necessary amount of data. The second one concerns the size of the neural network — the bigger the neural network the higher performance. 
AI in today’s business world
After establishing what AI really is all about, let’s take a look at its application in today’s business environment. Perhaps you would think that only big companies with a large budget and workforce can benefit from AI. However, AI can also be employed in small businesses. An interesting example offer Fujitsu and Microsoft: both have been working with Japanese dairy farmers to find out exactly when cows are in their estrous cycle in order to optimize artificial insemination. The insemination process is a very tricky task for farmers, as a slight miscalculation in timing can result in failure and delay the process for another month. After incorporating the farmers’ knowledge about the increased movement of cows during their estrous cycle into a systematic data analysis (i.e. by fusing it with AI), the success rate increased to 95%.
In spite of such success stories, the use of AI in businesses is not exactly magic. There are undeniably much more sophisticated cases, but usually “AI runs on data, companies need to know what kind of data they have, what data they have access to or what other data pool they can merge with their data and then they can reason across and surf the insides” .Thus, ”to incorporate AI to business strategy requires visionary leaderships in the company to recognize values of AI and to find out where business value is and what’s hard to copy” . Granted, companies cannot have AI teams for all their units as it would be unwise to build everything in-house if it is not very business-specific. This is why the acquirement of solutions that are widely used in the industry should be considered. 
JANZZ.technology provides both smart HR solutions for businesses and labor market solutions for public employment services. Since JANZZ’ establishment in 2009, we have been aware of the two decisive factors that ensure a precise performance of supervised machine and deep learning. For the past 9 years, JANZZ.technology has been continuously building the largest occupation-related neural data network in the world, with 300 million neutrals up to date. Our unique algorithm has been trained with large amounts of data from various partners in international corporations and public employment services, hence we are able to offer our services in 40 languages. We strongly advise customers to collect structured and effective data in order to get the smartest matching results.
Write now to email@example.com
 ZHOU, Chaochen. 2018. Gei Ren Gong Zhi Neng Que mei. URL: https://www.huxiu.com/article/247345.html [2018.10.31.].
 New Work Summit. 2018. Power Lines with Andrew Ng C.E.O. and Founder, Landing.ai; and Adjunct Professor, Stanford University [Video file]. URL: https://www.newworksummit.com/nws2018/gallery [2018.10.31.].
 New Work Summit. 2018. The AI Accelerater with Peggy Johnson, E.V.P. of Business Development, Microsoft and Frank Chen, Partner, Andreessen Horowitz [Video file]. URL: https://www.newworksummit.com/nws2018/gallery [2018.10.31.].
 Andrew Ng. 2016. What artificial intelligence can and can’t do right now. URL: https://hbr.org/2016/11/what-artificial-intelligence-can-and-cant-do-right-now [2018.10.31.].