There’s a revolution going on… We are at the very beginning and it is mostly described in terms of sensational news. But it’s serious.
How can you know that this text is written by an actual human being? Well, you can’t know for sure but you can however check my profile and find out that I’m with Vivaldi since the very beginning in 2013. That’s more than 9 years ago. At that time the artificial intelligence (AI) was probably on training so as to be able to do the things that we are witnessing today (and very publicly for a couple of weeks). However, the beginning of my membership can eventually be hacked, leaving no way to verify my identity. Now the AI writes texts, paints, creates art, poetry, music… faster and soon better than any human being. Writing is dead, music is dead, image-editing is dead. It’s a matter of time before the original content creators we thought to be invincible will be completely replaced… and even the most skillful critic won’t know the difference between the work of a human and the work of a machine.
What remains intangible for the AI for now? Intuitive stuff and emotional stuff. But make no mistake, this stuff can eventually be thoroughly defined, described and faked, if sufficient data is gathered. And gathering enough data is not a problem for the AI. Time is also not a problem given how fast the AI moves. This also shall be conquered. I expect to become irrelevant in my psy field in a couple of years, given the exponential development of the AI’s abilities.
However, I would like to point out some issues.
First, the accuracy and the predictions of the AI is based on the precision of the data it is trained on. To put it simple, it depends on what you feed the AI with. The internet is a bad source of information, as it is distorted by the subjective opinions of people. The internet is not scientific data, the internet is not “the Truth”. And you must decide first what “the truth” is. The internet is made of the opinions of other people and what is subjective, personal, is also biased. There are out there several “truths” but not a single “truth” (unless we create an authoritarian society, which is not so impossible given the recent trends of compelled speech and exclusion of different kinds). The internet is a mirror for the collective unconscious of the population inhabiting it; if you think that the internet is a jungle, rest assured it is, because the human unconscious mind (and often the conscious as well) is deeply chaotic. A world ruled by the mob on the internet is likely to be an anarchy and end up in a nuclear war (which seems to be more and more the case). People have developed technically but not morally, and this is quite well known; for this reason there are still wars: there is no shortage of psychopaths who rule other individuals who need to be abused. And I suspect that even the creators of AI (the companies developing it) are also deeply psychopathic, as they are aiming financial gain with little effort and no consideration for the dangers and consequences of misused AI. Just think about the discovery of atomic energy that has been converted very fast in a weapon that was also deployed and killed people; the AI will follow the same path because history tends to repeat, as the human mind remains the same throughout centuries. In conclusion, if you train the AI on false information generated by deeply flawed individuals, you will rapidly get a bullshit: a hyper-intelligent AI with the morality of a psychopath. And then the Matrix, as seen in the eponymous movie, is just around the corner…
Second, I see a phenomenon for some time: people around me have difficulties writing in their own native language and expressing themselves orally. For instance, I see the already classical confusion between “you’re” and “your”, but I also see the lack of use of comma where it is supposed to (not) be used. At some point texts become unintelligible, hindering understanding and communication. And given the fact that the clarity of the written text is an exact mirror of the way one thinks, you can infer the chaos in one’s mind by looking at their texts. The second problem is what could be called functional mental retardation: people are normally intelligent but they don’t use their mind to articulate ideas or extract fluently the main idea from a given text or situation. When you spend your day looking at videos, images or play games, you lose the ability to express yourself with words, you become more or less illiterate. In the end, you resemble a retarded individual despite having a decent IQ. Add to this the fact that you don’t only need intelligence but also knowledge (that means reading books) to orient yourself in life, and we arrive to the larval existence of humans in the Matrix (the movie) who please themselves to be connected to a virtual reality where they are not triggered to do anything but to passively witness their life without any possibility (and desire, frankly) to oppose. The fields of embryo-like humans being cultivated (harvested) are a very graphic image of this idea of passively witnessing your own life, being a spectator and never an actor. In conclusion, the over-relying on this tempting AI-enhanced world risks to increase the occurrence of a phenomenon that is already present; I can’t emphasize enough how hard it is to see a patient, in my role as a psy, and find out that he/she can’t express themselves, can’t identify emotions and can’t have access to rational/logical thinking, lying in a deep immaturity that prevent them from doing anything with them and their problems. How to negotiate with such an individual who is at the emotional & intellectual level of an infant?
Third, returning to the problem of AI, a lot is being discussed about who is to be blamed for the disaster that is to come. I’d focus on solutions, as what has been done is already impossible to avoid and given the greed of the humankind, the AI will continue to be used and become better and better. One solution is to keep using one’s mind so as to avoid being trapped in a state of comfortable numbness (and dumbness). This implies pain and frustration. One needs to be open to “suffer” to read a book when an audiobook or a quick 15-minutes-summary of that book are available. One needs to frustrate oneself so as to remain fit for knowledge and intelligence. One needs to still do their math homework when there is an app available for solving the same math homework in minutes. At the same time, it appears to me that what is human-generated will tend to become a niche exactly because of its imperfections. An AI-generated-photo will be less expensive than a human-generated-photo, but this will remain true only for already-established artists who will gradually die and will never be replaced. Finally, the offline-time will become much more valuable for people focused on their privacy who want to escape the AI-generated online world, but with the proliferation of security cameras and location-tracking devices, this will be increasingly hard to get.
What are the limits of AI? As I said, with a lot of information available, everything and everyone can be anticipated. But the design of the human being is not only rational and not only conscious. As a psy I know that humans are unpredictable. There is a dark place in each of us, our so-called blindspot. There are things about us that are public (we know them and the others know them as well – our Persona), there are private things (our Secret), there are unconscious things that escape our control but are perceived by others (our Shadow) but there is also a fourth category of things that are both unknown to us and to others (this Blindspot… or should we call it Mystery?…). Because it is completely unknown to anyone, it is also beyond the reach of the AI, as it cannot be defined. This Blindspot can create surprises; it can also generate solutions the AI cannot generate and, in the eventuality of a disaster similar to the one in Matrix, it will be our only chance.
While working with predominantly rational people, and with those using a lot of defensive rationalizations and intellectualizations, along with pondering on my own life, I know that the rational solution to this life is, inevitably, the suicide. If you are purely rational about this bad joke called life, you will come to the rational conclusion that the best thing to do is to kill yourself immediately. The data-trained AI, regardless of the quality of this data, will inevitably come to this same conclusion: the best thing to do with humanity and life is to kill it. Why? Because it is using reason, logical chains of command lines. It takes more than logic to create a civilization and a biological race. People have tried to describe humanity, quite atheistically, in the terms of a biological mechanism. This is probably possible for animals, whose reactions and decisions can be anticipated if one knows their instinctual drives. But for humans it is much more difficult because humans are not only conscious, evolved animals, but also have a transcendental (mystical? spiritual?) quality that cannot be fully described using a logical language. The AI, the machines, will always be able to replicate, execute, assist, do everything better 1000 times faster, but they are doomed to commit suicide and fade away, possibly taking us with them, if left to their exclusively rational decisions.