From telephone ladies to elevator operators to actors, we talk about the rise of artificial intelligence (AI) with Matej Viktor Tamás, an AI student at the University of Oxford. According to Tamas, the real AI revolution will happen in the background over the next two years, but it will also involve big changes.
Matej Victor Tamás is an asset manager at Hold Fund Management and a student of artificial intelligence at the University of Oxford. This is the fourth year that Oxford has offered AI as a one-and-a-half-year postgraduate course, and Thomas is the first Hungarian to study here. There are approximately fifty people on the course, and his colleagues include, for example, a security director at Netflix, a project manager at Amazon, and a data management director at Bloomberg and Vodafone. Training focuses on business applications of AI, with topics such as “Big Data and Machine Learning”, “Basics of Statistical Models”, “Deep Learning and Neural Networks” or “AI Ethics”.
Is the cheese evenly distributed on the pizza?
Tamás has been working at Hold Fund Management for the 16th year, where he is responsible for an investment portfolio worth HUF 60 billion. As a financial professional, he believes that AI will also rewrite the investment decision-making process, as it will be possible to summarize, clarify and simplify data faster and more efficiently. It will also be very easy to overcome language barriers, for example, to analyze a Turkish company, it will be necessary to know much less Turkish.
On the other hand, customer service, AI has really evolved and there seems to be a lot of pressure on everyone to start engaging with it. But here we appear to be far from perfect, at least in the highly regulated financial sphere.
“My bank introduced their AI assistant to their mobile app a few days ago, and I was surprised by the inability to answer any of my questions.
“Whether it’s the euro-forint exchange rate or the amount of money I have in my account,” Tamas explained. Because MI will definitely answer if you ask. “Which is, for example, what banks can’t afford – and clearly can’t afford – to answer inaccurately. That’s why I can ask my bank’s chatbot almost anything, and it gives an evasive answer. Because “You probably won’t be able to answer with 100 percent accuracy.”
We’re still a long way from achieving the high-quality conversation in finance that we can enjoy in our own lives using AI chatbots, but we’re getting there.
There is already some type of AI-based software behind many of the services we use. In fact, it has already taken over our lives in many areas.
- Uber, the world’s largest taxi company, doesn’t have a single taxi, it just has a software program based on artificial intelligence.
- The biggest media companies in the world are Facebook and Google – they don’t produce media content, they just have the right programming background. At Google, a machine learning-based mechanism helps us get the right result in an online search. Instagram, Facebook and TikTok constantly monitor how long we look at a particular photo or video. Then, by offering more and more new content, it more precisely filters the topic that interests the user.
- The world’s largest accommodation provider, Airbnb – and it doesn’t have a single accommodation, just an enhanced AI-driven software
– Lists examples
Tamas says he agrees with Andrew Ng of Google Brain (Google’s former artificial intelligence research group). – Mr. Dr.) with one of its founders, who says that most of the value growth that can be measured in money in AI will happen in the background in the next year or two. Moreover, with the company’s solutions that would have cost a lot to develop until now, while large language programs now aid in efficient programming, models that previously could have been made in a year are now available within a week.
For example, a large pizza chain specifically inquired of Andrew Ng about whether it would be possible to create an algorithm that monitors pizza preparation with cameras and ensures cheese is evenly sprinkled on the pizza. “It’s useful for them, but it’s not so useful that it was worth doing even with the technology that existed a year ago. Now after a week of programming and coordinating some surveillance cameras already in the kitchen, it’s really worth creating this solution.”
Within twenty years, he will be doing half the work
In a study published by McKinsey this summer, it concluded that there will be four areas in the first round where much of the work will be taken over or facilitated by AI.
- Marketing: In addition to creating texts, creating images and videos
- Customer Relations: More and more chatbots and phone assistants are being activated by different service providers
- research and development
According to McKinsey’s general forecast, artificial intelligence will take more than 50% of humanity’s work within 20 years.
Tamas tried to put the proportions that initially seemed intimidating into context. He explained: Humanity has a lot of experience regarding what it is like when new technology eliminates jobs.
“For our generation, it is normal for a driver to sit in a taxi, but not in an elevator. Two generations before us, it was normal for an elevator operator to sit in an elevator, as a taxi driver would sit. While it would probably be clear to my grandson that there is no driver In the elevator or taxi.
Change or a new situation is always normal for the next generation, and more difficult for those who experience it.
A hundred years ago, the Bell Telephone Company was the largest company employing hundreds of thousands of women in the United States, so many operators were needed. But fifty years ago, this profession no longer existed. After researching the afterlife of former call center workers, they discovered that although many of them had not found a job as good and enjoyable as in a call center, the next generation, and their children, could choose from much better jobs thanks to the automation of something. It was very manual. “It takes time for people to transition and adapt. Incidentally, even though telephone work has ended, the most conservative corporate and office managers still ask their secretaries to call them and connect them with the office.
With the AI revolution also new jobs have emerged: for example, real-time engineers and ChatGPT consultants, that is, all jobs related to how to teach a large language model as accurately as possible. On the one hand, this is necessary so that someone can advise users on how to write requests and instructions that make their tasks significantly easier and shorter. On the other hand, developers of large language models have to hire more and more people to filter out hallucinations in the answers. Tech giants are working a lot to make the answers as accurate as possible, so there will be progress in this area as well, but according to Tamas, human contribution will still be needed for truly accurate and high-quality work.
What is certain: to achieve the same result, humanity will need less work overall. “You can be sad about what is happening to jobs. There will certainly be some very unfortunate stories, but on the whole, it is better for humanity to live a more comfortable and resourceful life.” Those who fear that people simply won’t find work anymore point out that many “It is often necessary to introduce a basic income in connection with this.” This topic goes too far. “Development in itself is a good thing, especially if it is carried out in a sustainable way, with better care for the environment.” As an example, he mentioned the route optimization carried out by various transport companies, which saves not only fuel but also emissions.
Big mistake from the financial sector
Everyone in the financial sector is feverishly developing AI advisors to help clients. However, according to Tamas, this is a big mistake on the part of the financial sector, because in three to five years, people will get used to having a global advisor on their phone, calling him up and talking to him all day long.
“They will ask him or her to book a restaurant for the evening, organize a meeting, or substitute there. These personal virtual assistants can not only listen in and summarize a Zoom or Teams meeting, but they can also speak in place of the person during negotiations. “People will get used to having their all-in-one personal assistant on their mobile phone to manage matters throughout the day, including their investments and banking. I think banks will be able to use AI advisors for a few more years, and then they should be ready for customers to let their personal assistants carry out various transactions.
Film is another area where, according to Tamas, big changes can be expected, and even the entire voice acting profession may become completely redundant. In the first place, actors do not have to fear that their voices will be used without their permission and that the voiceovers will be generated by a machine. The problem for them would be, for example, that Pierce Brosnan would gladly agree to have his voice heard in all the languages of the world. “This will be done in full agreement with the studio and the actor. There will be no need for dubbing, or at all.” Something similar could happen with podcasts: Spotify recently came up with an initiative to start streaming the conversations of the most famous American podcast hosts in their voices, but in other languages.
End of parties and phone hung up?
At the same time, the connection between things and artificial intelligence is becoming increasingly popular. Ray is a manufacturer of sunglasses, for example Collaborated with Meta: In collaboration with the parent company of Facebook, they recently launched a pair of glasses containing artificial intelligence, which can be used to take a camera, live stream what we see, and even speak to our ears if we ask it for something.
“I’m impatiently waiting to finally see ten thousand phones hanging up at concerts, because everyone is using them to take a camera or take pictures, but only the camera in the glasses has to be turned on.”
Also, not so long ago, there were also gadgets that could be hung around the neck or attached to clothes, which contained artificial intelligence. For example Eavesdropping and recording necklace, who also takes notes and summaries of what was heard. Of course it raises a lot of questions about what and who you are allowed to hear and what you are not allowed to hear, because of course there are things that require prior permission from others, but the point is that you can customize the device well.
Tamas mentions as another example that if we enter a hotel, it will be normal for there to be a display screen, or a tablet, where a nice smiling face welcomes us, speaking to us in our native language and telling us what time breakfast is served, or when we can use the swimming pool. “Many people have language barriers that prevent them from traveling comfortably, because they do not speak English well enough to be able to clarify more complex matters with hosts and hotel workers. AI will address this very soon as well.
Cover image: Laszlo Sebastian