On 25.07.2022, Levente Szabados, co-founder and senior advisor of Neuron Solutions, was invited to philosophize again in the Millásreggeli radio show. While the previous month they discussed the possibility of the computer’s awakening to consciousness, this time the radio hosts discussed the actuality of artificial intelligence with Levente. You can access the previous interview at this link.
It already makes us smile to ask “How far away are we from using AI systems?”, because AI is part of our everyday life. But do we have a good understanding of what lies behind this nebulous concept that is an integral part of our lives?
Levente believes there is a common misunderstanding on this topic, namely the confusion between the concepts of intelligence and self-awareness. In fact, at least according to the research, we can probably agree on this: “Intelligent is something that solves a task (whatever it may be) well.” “In return, it may not have a very rich, subjective experience.” So that’s not enough to talk about consciousness, yet in people’s minds some sort of know-it-all intelligent thinking computer appears at the mention of the magic word artificial intelligence. But that is not the case.
So where does AI stand today?
Well, as we said, we are talking about machines that can only solve tasks. Tasks like goal-oriented pattern recognition, personalised product recommendation, and an endless list of other tasks. The uptake of AI is astonishing, even Levente was amazed at the numbers he brought to the show: according to an IBM survey, 35% of companies surveyed are already using some kind of machine learning system, and a further 42% are experimenting.
So the current state of AI is that there’s a lot of experimentation going on, but mostly on small target tasks, each with a different method. However, there is research into creating machine learning systems that already attempt to handle several types of tasks (e.g. visual and textual) together. Experiments to create such more broadly applicable models – fundamental models – have been going on for about 1-2 years, so it is still in a very rudimentary phase, but it does present a very interesting sci-fi picture: Will we succeed in inventing a machine learning system so that the same model can be used for, say, piloting an airplane, classifying literature and describing images? Now that’s a matter for the future.
This topic is too interesting to end questions. Levente also gives his opinion on the question of when the age of cognitive artificial intelligence will come, i.e. an age where machines are capable of a very wide range of tasks, where they can answer and act on any question or request we make.
“The future is here, it’s just not evenly distributed.” Separately, professional models already exist. There are models that perform mechanical tasks, that can turn the heating on for us at the time we want, and anecdotal models, see previous discussion of Google’s artificial intelligence. However, the two are not yet connected.
I suppose you have also asked yourself: was Levente referring to the anecdotal robot vacuum cleaner? Well, maybe the next conversation will reveal that. We’ll be back in a few weeks with the next episode, so follow us and don’t miss it!
You can listen to the full conversation in Hungarian in the following Youtube video: