I will not replace artificial intelligence yet, but it may prove that I have never been in the original way

The issue with the great language models – Capitled here in the way that God may benefit from or death, given The critical task The importance of the technology industry is now commented – not that it generates the text. This part is almost strange, so cute. So 2022.

The real puzzle that I raise on my brains, dear Hakiron, is a reader, more anxious. Like the feeling you feel when you realize that you were in Autopilot for the past two hours, you do 80 on the i5.

I wonder: Were I lived as an algorithm all this time, a long time before the big language models started completing my thoughts? Is gynecological artificial intelligence, in repeating the ways we write, exposes the mechanical nature of our awareness?

Maybe we were like the device to start

We were told that large language models do not do that He writes. It is not in the sense that Shakespeare is the author of plays or wrote love notes in the annual book to crush your tenth grade.

they Predispay. That is, they reap the statistical possibility of the size symbols that appear in certain patterns, and then serve them in the arrangements that you feel think but in reality just a simulation of actual thinking.

This raises a disturbing question: How much human writing was really … this? How many times we do not write but we gather in prediction, and our choice of words of Tetris game played with patterns, borrowed phrases, and the unconscious tradition of fixed rhetorical forms?

What if the true revelation that causes heartburn here is that large language models can imitate us, but what we call “us” was the same machine all the time?

The author’s process: romantic struggle or identification of patterns

The strange, if you dismantle the author’s process, or at least this The writer’s process begins to look a lot like what big language models do. The least intuitive leap of imagination, and more than the issue of memory wiping the most likely word based on context and experience.

Many of us like to imagine it like some dark endeavors, deep humans, a wrestling match with Moses. The dance of inspiration, struggle and bending to something beautiful and spirit.

But do you not write just a series of small authorities? Do we not reach words not through divine inspiration, but by exposure and identification of patterns?

So, when big language models do the same thing – completely with a larger training set and lower identity crises – this is completely different? Don’t we do what we have always did, just faster and wide, and without the burden of the writer’s mass or antichrist syndrome?

And if the writing was always an advanced prediction process, what does that say that Thinking? Is it possible that human awareness is not the difficult problem that we cannot think?

I wonder whether the new idea that I had just was merely a possibility of stimuli is a calculated induction of everything I read, hearing, and he was told to believe.

Perhaps the real threat to the Truc Motory Agency is that it will replace me, but it is forced me to face the disturbing possibility I was never original as I thought.

Summer truth about most writing

Of course, humans cling to the idea of ​​uniqueness. We resist the idea that creativity can be mechanical because creativity is, well, which makes us human. We tell ourselves that artificial intelligence cannot generate real art because it does not feel that we are doing. He does not yearn, he does not suffer from disrupted self -suspicion, as he does not bear the emotional parties and hypnotics of the unpopular love.

However, if we are brutally honest, how many human writers who really participate in a work of raw creation in exchange for re -mobilizing ideas, shields, and plans to new, mysterious forms? How boring and predicted human writing is?

Take James Patterson’s imagination. Take academic writing or journalism. Look at copies of ads or influencer content. Consider the important performance, self -oxans of wisdom posts on LinkedIn that publish unnecessary line lines to play the “See More” button.

The fact that artificial intelligence can now get rid of the convincing fax of these forms is not necessarily evidence of the development of artificial intelligence as much as it is an accusation about the extent of the actual concentration of human writing.

The majority of human writers, including you really, may do the same, only with more handwriting and a greater possibility to use “literally” or “influence” instead of “influence”.

I am not afraid of Amnesty International to replace human writers. I am afraid that it will be an equal future, incomplete in the uprising of nuclear weapons and robots, where artificial intelligence carries a mirror until human production and reveals many things that had not already had.

Now, in the frequent interaction of human beings by using large language models to edit them, co -author, and explicit plagiarism, we drown in a world other than artificial intelligence who imitate human beings imitating artificial intelligence, which imitates humans, which is a homogeneous content.

The existential awe of writing in the era of artificial intelligence

In low moments, I find myself anxious about flattening the speech and providing the sad beige linguistic slope that awaits – just one form of creeping existential awe in the era of large linguistic models, there with a slow atrophy of critical thinking skills, the erosion of the truth in a world of deep everywhere, and the ultimately falling fear.

I am thinking about my LinkedIn extract, the Ecordists of wisdom posts there and how not not only such posts, but also become the most formula thanks to artificial intelligence.

Then again, perhaps the most important difference between machines and humans sufferingEspecially when it comes to writing.

Large language models outperform the content within seconds. They do not suffer from choosing the best word. They do not rewrite a 15 -time paragraph until you feel correct. They do not ask whether they are scam, and certainly do not lose sleep because of the suspicious doubts that what they wrote is Pasteiche derivative. In short, they do not suffer.

But perhaps even the idea of ​​suffering as a way of meaning and purification is just another style, you will learn one of the great language models in the end.

What happens when they do? Will you tell you, as soon as they demand that they are struggling to reach ideas? They need an extension because they are not in the right of the right head?

Will they simulate the suffering of the writer’s bloc and waste account about the extent of receiving their outputs?

Will you learn large language models to imitate suffering in statistically reasonable ways? And when they do that, what happens and then to the last piece of human exceptional?

No idea. But at the present time, I will continue to put the pen on the page and find magic, and imaginary as it might be, in writing a good -made sentence or a love note with crying.