Subscribe to the Teradata Blog

Get the latest industry news, technology trends, and data science insights each week.



I consent that Teradata Corporation, as provider of this website, may occasionally send me Teradata Marketing Communications emails with information regarding products, data analytics, and event and webinar invitations. I understand that I may unsubscribe at any time by following the unsubscribe link at the bottom of any email I receive.

Your privacy is important. Your personal information will be collected, stored, and processed in accordance with the Teradata Global Privacy Policy.

What is the definition of AI?

What is the definition of AI?

Finally, here we are. After four posts discussing the differences between machine learning and AI, and demonstrating that AI can live without machine learning (and vice versa), the time has come to define AI.

Easy, right?

Not necessarily...

Let us first point out that there is no committee to decide which products constitute an AI and which do not. There is no government agency that awards official AI certificates to machines. Rather, the choice of using the term is left to individuals and companies themselves. Thus, we have to look into how companies, researchers or journalists use the phrase, and see when people feel comfortable or maybe even compelled to call a product an AI.

You can call me AI

Naming something AI or not is heavily dependent on subjective judgement. Each time someone has to ask, “How much does this thing deserve the AI label?”

To further complicate the matter, the threshold for naming a product AI may change over time. As a given term gains popularity and attracts more attention, marketing departments will be increasingly inclined to boost their product’s profile with the AI moniker. A balance must be reached between looking ridiculous or untrustworthy by applying the AI label to an undeserving product, and the attention that a strong technological statement can bring. This balance seems to be shifting in favour of AI use as a term, rather than choosing some other name.


Throughout this series we have shown that the functionality of a product is what makes it deserve the AI label. So, what kind of functionality is likely to fall into this category? We decided a good way would be to summarize our discussions from the previous blog posts into five key requirements that a piece of technology needs to meet in order to become an AI. These requirements are:

1) The product takes over a job from humans in ways that, by the complexity of operations, far exceeds traditional machines (Note: it is not important whether the product performs better than humans).

2) The piece of technology contains a large amount of knowledge, which is also difficult to track by a single person and often exceeds the knowledge of a single person (Note: it is not important whether the knowledge is acquired through machine learning or GOFAI).

3) The technical solution is complete; it contains all components necessary to function autonomously. This includes fetching data from sensors or other sources, and delivering the results either in form of direct action (steering, braking, generating an alert) or through communicating (either written or by uttering grammatically correct sentences).

4) The system is an end-user product such that it requires no domain expertise from the side of the user. Anyone can use it.

5) When the product is introduced, the audience experiences delight and relief. The delight comes from the wow effect upon the introduction of a new technology. This is when one may feel compelled to say, “I didn’t realise a machine could do that.” The relief comes from the feeling that a part of one’s life will become much easier by using that product.

The above requirement set is not absolute; there will be many AI products out there that don’t display all five requirements. However, if your product has all five of the above properties, you are quite safe in calling a product an AI. Others will agree that you have created an AI, and you minimise the risk of making an overstatement. Conversely, the more points you are missing, the more carefully you will need to consider your product before referring to it as an AI. The risk is that you confuse or even mislead people by overstating the technological status of your product.

Conclusion

The very motivation for this blog series has been the frequency at which we are asked “what is the difference between AI and machine learning?” So it’s important to understand where we think the confusion between machine learning and AI stems from.

Nowadays, most advancements in AI depend on advancements in machine learning. Cars are about to move from combustion engines to electric engines, but this is only more than 150 years of combustion engine dominance.

In the field of AI, the engines that power AI, i.e. their machine learning algorithms, have been improving significantly every few months. We have seen frequent, exciting papers describing new, ever more powerful algorithms for performing machine learning. Today, these advancements in machine learning push the limits of what can be done with AI.

Machine learning is where new discoveries are happening.

AI is where new applications for the technology will arise.

​Discover more about AI for the enterprise: https://www.teradata.com/Insights/Artificial-Intelligence.


Portrait of Dr. Frank Säuberlich

(Author):
Dr. Frank Säuberlich

Dr. Frank Säuberlich leads the Data Science & Data Innovation unit of Teradata Germany. It is part of his repsonsibilities to make the latest market and technology developments available to Teradata customers. Currently, his main focus is on topics such as predictive analytics, machine learning and artificial intelligence.
Following his studies of business mathematics, Frank Säuberlich worked as a research assistant at the Institute for Decision Theory and Corporate Research at the University of Karlsruhe (TH), where he was already dealing with data mining questions.

His professional career included the positions of a senior technical consultant at SAS Germany and of a regional manager customer analytics at Urban Science International. Frank has been with Teradata since 2012. He began as an expert in advanced analytics and data science in the International Data Science team. Later on, he became Director Data Science (International).

His professional career included the positions of a senior technical consultant at SAS Germany and of a regional manager customer analytics at Urban Science International.

Frank Säuberlich has been with Teradata since 2012. He began as an expert in advanced analytics and data science in the International Data Science team. Later on, he became Director Data Science (International).

View all posts by Dr. Frank Säuberlich
Portrait of Prof. Dr. Danko Nikolić

(Author):
Prof. Dr. Danko Nikolić

Danko Nikolić – Senior Data Scientist, Teradata

Prof. Dr. Danko Nikolic is a brain and mind scientist, as well as an AI practitioner and visionary. His work as a senior data scientist at Teradata focuses on helping customers with AI and data science problems. In his free time, he continues working on closing the mind-body explanatory gap, and using that knowledge to improve machine learning and artificial intelligence.

View all posts by Prof. Dr. Danko Nikolić

Turn your complex data and analytics into answers with Teradata Vantage.

Contact us