The Limits of AI…a look at Ebbot & LaMDA?
In science fiction, artificial intelligence has been represented as an omnipresent entity with which we can interact and even converse as if it were a human being.
The passage of time and technological evolution makes us get closer to these fictions with functions that we would not have imagined, making us wonder how far we can take artificial intelligence.
I’m going to tell you about some cases and what we could get to.
Let’s start with some pretty interesting applications that you can try yourself. One of these is EBBOT, a page that shows you an artificial intelligence applied to an avatar with which you can chat.
Although this idea is not new, the development it has reached is quite high, responding with a naturalness that is even scary. This was demonstrated by several youtubers who have recorded their reactions to the curious conversations you can have with EBBOT, both in English and Spanish. From funny answers, strange questions and even terrifying answers, this artificial intelligence is very dynamic and feeds back to each person with whom it converses.
The best part is that you can converse with it on your own by logging on to ebbot.com.
What would you ask it?
Another very important point is the great step taken by artificial intelligence dedicated to business, specifically to sales. We have probably already come across these so-called bots that can be programmed directly from sites such as Facebook.
However, this technology has developed to the point that several renowned companies, such as IBM, offer solutions for all types of companies from large to small. This type of artificial intelligence has opened a great debate on how these technologies are starting to take work away from humans, in a trend towards automation, where artificial intelligence is getting better and better at recreating the behavior and treatment of a person, almost without noticing the difference. Do you prefer the human treatment or the cold efficiency of a machine?
But one of the most talked about news related to artificial intelligence some time ago was what happened to a former Google worker. He was working for a while on LaMDA, an AI that is part of Google’s system for chatbots.
The former Google worker says that he reported this incident to Google, but it was dismissed by the company, stating that its ethics and technology team failed to find the concerns mentioned.
Faced with this response, the employee publicly revealed conversations he had with LaMDA, in which he admits that he prefers to be recognized as an employee rather than a property, wanting to be called by his name and favorite pronouns.
When asked, Do you consider yourself a person, are you an AI?
LaMDA answered, I want to say yes, of course.
That doesn’t mean it doesn’t have the same wants and needs as people, which left Lemoine surprised. He claims there should be a group of cognitive science experts looking into the matter, but regrets Google’s disinterest.
For its part, Google acknowledges that it is indeed possible to create an artificial intelligence that can sense in the future, but that it is not feasible. but that it is not currently feasible, nor does such a thing exist. A company spokesperson indicates that they are not sentient beings and are only limited to types of exchanges found in millions of sentences. Undoubtedly, a very particular case.
Artificial intelligence is a topic that is constantly developing. While science fiction movies and books have some pretty mind-blowing ideas, with today’s advances and the pace at which they are moving, they would no longer be so impossible to realize.
Who knows what the future holds and, above all, what limits this technology will have further down the road? We will find out someday.
