In 1950, an artificial intelligence was a machine capable of simulating human thought. It has taken us some time to realize that this task must be “cut up” into smaller ones. That is why today we are more modest, and we are satisfied with an AI successfully simulating a small percentage of ourselves. But what will become of artificial intelligence in 2025?
There is no doubt that, as it progresses – by leaps and bounds, as it seems – we will also change what we understand by it. Artificial intelligence currently helps us to optimize processes. Given its versatility, in the future we could frame it in many other areas.
The IA will be in charge of services that we don’t even think about.
In very general terms, technology related to the “thinking” of machines is related to delegating tasks. First, the calculation, then the predictive. It is possible that in 2025 we will leave everything else to them and allow them to make decisions for us if we discover that it benefits us or improves our quality of life.
It is very likely that at some point in the next decade we will delegate much of the work conversations to machines. Today technology already allows us to call a local for us and reserve a table (above). We will probably give the machines a certain degree of social solvency, “letting go of the belt” with which we now tie them.
As we saw at the time, consultants such as PwC and Gartner believe very likely that smartphones with artificial intelligence will become a digital “I” of the user, a sort of virtual butler in the style of Iron Man’s JARVIS. As machines become more intelligent, we can delegate more tasks. Also, more responsibility.
Making AI available to all
Statistically, we’re unlikely to know how to program a neural network. Most of us don’t have this knowledge, just as we don’t light a fire when we want to warm up. Instead, we turn on the heating. It seems that the near future of artificial intelligence will inexorably pass through its democratization in all directions, perhaps starting with employees.
What the AI will bring us in 2019
High-tech companies have a bottleneck in the development department. They are the ones who invent the future every day, but there is a lot of labor friction when trying to transmit day by day part of this knowledge to other verticals of the company. Meanwhile, artificial intelligence is gaining skills.
We will see the real jump of AI when it can be programmed in a simple way by an average user, in the same way that he is able to investigate a little and start creating tables, dynamic tables or macros of increasing difficulty in Excel. The day when these tools are accessible to the average population will begin “the revolution”. And it will probably be soon.
The revolution has not yet begun
A year ago, Michael Irwin Jordan, one of the world’s undisputed leaders in machine learning, wrote an article entitled ‘Artificial Intelligence – The Revolution Has Not Yet Happened’. A few years earlier, Tim Urban of Wait But Why wrote about the “artificial intelligence revolution” and the road to super intelligence. The graph describes the phenomenon quite well.
It depicts the “edge” of human progress. Despite our efforts, it will look linear once the potential of AI and the emerging systems, whose unpredictability we discuss later, have been unleashed. To understand scale, we bring a riddle that is used in classes to teach the speed with which these emerging systems manifest.
“You are inside a barrel with a drop at the bottom. The amount of water doubles every minute and in one hour the barrel will be full. In what minute will you drown?”
Our brain is not developed to think about disruptive changes, but there is no doubt that these will come at some point. Almost certainly by 2025.
Emerging systems cannot be predicted
Who could have predicted the explosion of knowledge after the printing press, the automaton of the scribes? Something similar happened with smartphones. One, isolated, is a very useful pocket computer, but the emerging system that changed the world a few years ago came from millions of them interacting with each other. Its consequences could not be accurately predicted.
Now that devices such as the Huawei P30 Pro already incorporate chips oriented to artificial intelligence, a new emergency is being cooked (in the sense of emergent) whose end we cannot imagine. We smell something, but fixing it on paper won’t make it more likely.
Let us take reflection, for example, to autonomous mobility. It is estimated that there are more than 330,000 professional drivers in Spain. In a decade, as far as we know, a third of a million people (0.7%) may have to look for a new job. Imagine the impact this has on other sectors.
Now let’s think that there are already machines that make hamburgers, analyze the stock market, translate, organize trips, generate summaries and even paint pictures. The two best chess players in the world are not peopled, they are artificial intelligence. Slowly, machines learn to do something, and once they do, they don’t forget it.
We cannot predict the future, and there is the paradox that the more data we have to try to analyze it, the more divergent the approximations seem. But we do know that there will be changes. Soon a new emerging system will emerge, perhaps educational or economic, and the key will be to understand what is happening and how to adapt to change.
A unitary AI made up of many specific ones
We opened this article by saying that in 1950 machines were going to be as intelligent as people. Alan Turing fantasized about making them indistinguishable from us. Along the way we learned that it was easier to create artificial intelligence for each specific problem. The question we can ask ourselves is: can we unite them into a larger AI?
Perhaps by 2025 we will use a succession of artificial intelligence coordinated with each other by natural language processing. Interconnected, from our limited human perspective we will seem to speak with a single assistant while, behind the scenes, hundreds or thousands of “small” AI will solve each demand.
According to all experts, we are at an early stage in the development of artificial intelligence. In a decade, the breakthroughs we are astonished at today will go completely unnoticed. We are living in an exciting time, and soon we will have to redefine what artificial intelligence is to us.