Artificial intelligence is yet to have an unprecedented transformative effect
Although the term artificial intelligence (AI) has been around since the first computers first started to resonate with the professional community, it is only today gaining the most fame. Thanks to the advent of Chat GPT, its influence is also permeating the mainstream internet, and AI is gradually coming to the forefront of interest and discussions that have long since transcended the boundaries of the technical community.
How does LUKÁŠ MATĚJKA, founder and leader of the technology startup Zoe.ai, perceive this hype around AI? The recommendation tool Zoe.ai has been using the principles of artificial intelligence and deep machine learning for several years and contributes to optimizing business and delivering tailored offers to specific users.
Most tech companies are trying to implement AI in their own way. At Lundegaard, are you sticking to what you already know, or are you finding the Chat GPT model and its features?
At Zoe.ai we have been using AI and machine learning algorithms for some time. The algorithms look for user behavior patterns and help recommend the best within the offering. As far as Chat GPT is concerned, there is a general language model that answers specific questions. On the other hand, we use the language clients speak on the web. In the world we operate in, no general recommendation model has been created, trained, and provided to the whole world. We don't have anything like that yet. That's the noticeable difference.
We could not use Chat GPT for one simple reason. We would say to the tool: now I'm going to go shopping on this e-shop, and I want to recommend something to buy - the GPT would not recommend anything to me as it knows nothing about me and doesn't know what I've done or is doing there... I would have to tell it everything myself first. I would have to tell it what the e-shop sells and consider 10 other things to make the recommendation workable.
So basically the current hype around Chat GPT hasn't brought anything groundbreaking for your team?
Yes, that's exactly right. The principles that GPT Chat uses have been introduced previously. They've been around for a while. The revolutionary is that it's such a massive model, trained on massive amounts of data, and it can do things that the creators maybe didn't expect it to do. Plus, it's broken into the mainstream in a big way, but the techniques have been around for a while, and it's not a dizzying technological breakthrough. And I would just point out that Czech Tomas Mikolov brought the coding language techniques that are the basis of today's models. We have that Cimrmanian footprint in Chat GPT that nobody in the world knows about. And that is beautiful.
In discussions, the boom of AI and ChatGPT are often dropped concerning the fact that they will one day replace programmers and other professions. Could this happen in the IT industry as well? Will it move in that direction, and will there be any fundamental transformation of society?
Absolutely, I have been saying for several years now that AI has had a transformative effect on the economy, society, and culture and will have a much more significant effect, and that this has become a mass reality since 5 December last year. I totally agree with that.
A huge number of projects have been getting off the ground over the last six months to a year. These things and changes will likely keep happening over time. There is much scaremongering in the mass media, instead of saying that this is just another tool that humanity should learn to use and actually enrich itself with. Like everyone uses the internet today, it's not a specific advantage for the average person. Instead, many new professions will emerge, and the world will have to adapt. For example, graphic designers won't be creating 4 designs all day, but they will complete maybe 1000 designs in a standard shift because they will be able to help themselves. Those who don't adapt may lose their jobs, but Chat GPT won't start to be some extreme advantage, which it may be for some now, but in the final year, it will be an everyday thing. It will have a huge transformative effect.
What are the current pitfalls of using ChatGPT?
Copyright and security are huge issues. There has been a massive series of lawsuits in the US and Europe against these types of tools that have learned their art from nothing more than so-called ripping off other people's content. Part of the obstacle is that one cannot be sure whether the AI is wrong and whether one can trust its factual knowledge 100%. It is not connected to the internet and up-to-date information, so there is still much work to be done to mark which data is verified and which is not.
There is also the problem of how to regulate this. Who will direct the answers? Will it be the government, the city, or the company? These are issues that are not addressed at all today and need to be addressed to make the whole thing work more reliably.
Will AI development and art go anywhere? Where? Will the hype around it fade or will the response remain the same?
If we focus on what has happened from December 5 to now, it's a really brutal pace. The number of language models that have been created since the end of the year, the number of add-ons and the ability to leverage other resources, and how many companies are in it - it's really exponential growth. The pace of things and the AI trend will accelerate even more. Just to read the news about AI is not even in human power anymore. Eventually, we'll have to use AI to summarize new information for us in that technical environment because the human brain won't be up to it. So yes, it will shift, and evolution will accelerate.
Isn't it also possible to touch on moral issues and pitfalls? Can AI be seen as a "rip-off"?
Humans don't work any other way. We take in information from all directions and make predictions based on it. Although in this respect, the problem of copyright is eliminated, on the other hand, the average person functions similarly and receives information from his environment, which he then works with.
What about attributing human characteristics to artificial intelligence?
Mentioning human characteristics in the context of AI bothers me. Many people introduce things into this topic that I don't see - in the sense of esotericism and the acquisition of human qualities. It doesn't have them, and it functions entirely differently. I like the analogy of her being like an alien intelligence. There are factors that make up certain characteristics, but they are not human. There is no consciousness or morality.
Lukáš has more than 20 years of experience in technical and management positions. He programmed one of the first "big-data" web archiving applications in the Czech Republic, implemented large-scale applications for IBM, was responsible for delivering web applications and portals for large companies, and led developers at Avast. In the last 5 years, he has been working on artificial intelligence and Fast Data technologies. With his team, he developed Zoe.ai, a tool that processes and interprets data from online behaviour in real-time and, for example, recommends tailored content to specific users.