On September 14th, co-founder and the Chief Product Officer in Mechanica AI Elena Samuylova gave a lecture about artificial intelligence (AI) in business. Elena told students about the differences between machine learning and AI, new market of solutions based on machine learning and its benefits for business..
Artificial intelligence is the ability of computers to perform complex tasks and exhibit human like intelligence. “Google is AI, Uber is AI, Netflix is AI. You are using recommendations of the movies that are shown to you, this is AI. So, actually, I’m absolutely sure that all of you are using it, just kind of inadvertently, not paying attention and not recognizing that this is AI,” — explained Elena.
There are two kinds of AI: strong (general) and weak (narrow). Strong AI you can see in the movies like “Her” or ‘The Terminator”. This kind of AI doesn’t exist in reality and unlikely to appear in the near term. What does exist is a weak AI. It is the ability of a computer to perform a specific intellectual task.
Today artificial intelligence is based on machine learning. This two terms are often used interchangeably. Machine learning is the ability of a computer to make decisions, to take actions without being explicitly programmed to do these specific things. According to Elena Samuylova, machine learning is based on three components: mathematics, data availability and computing power.
Often in connection with the concept of "artificial intelligence" people and media mention neural networks or deep learning. However this is just one of the machine learning algorithms that exists and it isn’t the best one. According to Elena, for some tasks it is good and for others it is bad. “So, honestly, you should forget about neural networks. Seriously, don’t think about it as ultimately the most important thing. Think about this general concept of machine learning and how this can be applied to different problems in business,”— Elena advised and told students about businesses that are already using AI and will be able to use it in the future.
First of all, AI can become an enabler of specific new products, new services, new applications that didn’t exist before and could not exist without AI. For example: self-driving cars and virtual assistants.
Secondly, “AI as a building block”. Artificial intelligence could be used as an important but not unique component of new decisions and business-models. For example:
- As-a-service business model. Half a century ago Rolls Royce company used this model for the first time to sell its “power by the hour” engines and with the spread of data analysis technologies, the share of such business models will only increase.
- Disintermediation. For example, in the future metal producers would have some kind of their own Amazon and would be able to connect throughout the supply chain with their end customers, the providers of raw materials and so on.
- Mass customization. AI will provide the ability of the producer to customize each item based on the needs of the final customer. Using AI you can learn to automatically adjust the parameters of the machine more quickly and cheaper.
The third idea is to use AI to optimize the already existing process to make it better. This is what Elena’s company Mechanica AI is doing. Using machine learning it is possible to predict the quality of products, “virtually” measure the necessary parameters or more efficiently manage the production process. For example AI can help steel makers to decide how much of each expensive ingredient they should put during steel production process.
According to Elena, the AI market is now growing. There are already many companies that produce equipment and tools for data analytics, are engaged in the storage and processing of information and develop business solutions based on AI. At the same time, these companies need not only technical specialists, but also people who do marketing, sales, consulting, digital transformation of corporations or project management.
“Managing data science project is actually art in itself,”— said Elena Samuylova. According to her, the word “data science” has term “science” for a reason. “ Comparing to software development where you can more or less understand how long for example it will take to resolve some goal and you can at least assess that the implementation of specific project is visible with data science project it is different. You run this experiment, you see the results, this algorithm doesn’t work, maybe I should change the data, so this is an iterative process,”— explained Elena. Therefore, when a specialist in big data works in a large corporation, this company needs to learn how to interact with him, and the project manager is responsible for this work. “The fear that we will all lose our jobs because of AI is a bit exaggerated. It mostly comes not to the automation of jobs but the automation of tasks,”— said Elena. According to her, in the future there will be fewer professions related to routine actions, but there will be more jobs for skillful specialists.