From the Agricultural Revolution to the Digital Age, Language served us well. The smart use of AI technologies can lead us to where language restrictions cannot be obtained.
Riding language from agriculture to lighting
People's development is closely related to technology. First of all, we expanded what evolution provided us with the use of tools during a certain period. We used stones, sticks, shells to expand what our muscles, nails and teeth could achieve. But we were not unique in it. The tools are also used by other monkeys and some birds and marine creatures. Where we separated from the crowd, it was when we were able to hand over how to make better tools from knee to another. Collect improvements over the years.
To do this – we needed better communication tools and brain power. The important thing is that for about a million years – we used stones and pulps (“Stone Age”) and only about 40,000 years ago – we started agriculture. “Surprisingly” enough – it's about the same time as the modern language has matured and more complex concepts could be transmitted. Summary, future, conditions, metaphors and so on. Without them, it would be difficult to explain if Bratt stands next to you in the grass, how to choose a specific seed type would give better crops next season.
It worked quite well up to the Enlightenment and the Industrial Revolution. During these times, the seeds of fast technological and thought worlds made the hyperchart of human development. The easiest way to illustrate this is through population growth – a clear indicator of the success of the species.
During this period, we invented things with growing complexity – government, law, finance, medicine, science, mathematics, transport, communication tools and last time – calculation. Until 1600, it was considered a cool Renaissance era – architect, doctor, sculptor and politician – it would be one doctor I would prefer not to visit today. This technological exponent is the culmination of our mind and knowledge – language transfer tools. Multiplied with the growing population – we were able to invent and interact with smarter people – which makes smarter people and so on.
But at that time (in the 1600) people began to suspect that something was broken. René Descartes and John Locke thought it was inaccurate and interpreted, Rudolf Carnap argued more logic and Noam Chomsky admitted to generating the meaning of the meaning. Claude Shannon is the father of his information theory of information theory of the digital age and communication bandwidth, by placing a language doubt as an ineffective means of transmitting information. Modern cognitive science, AI, linguistics and philosophers work and hope that modern technologies will make a solution.
Why language affects life for the rest of the life (300 years ago) is serious about our modern life and delays in another article for humanity. Nor do we discuss proposals for the improvement of word -based languages, compaction, non -verbal media and fidelity. The following is a look at the current state and music of whether the freedom of language pounds can be looked at on the horizon.
Foundation models and ants
If we talk to Chatgpt, twins or Claude (you thought correctly, named the above, Claude Shannon), we ask us to have a foundation model. All of these companies took a huge amount of text from everywhere that was able to put their hands on – every page, books, articles, code – and let gazillion on computers run LLMs to form a unit that holds the relationships and words relationships with each other.
It is a bit difficult to visualize, so we try the metaphor.
Niche of some ants sculptures. Think about the foundation model like a huge, complex ants. Each tunnel and chamber denote the relationship between words that have algorithms instead of ants. The deeper and more complex the ants, the better the model understands the nuances of the language and the relationship between the concepts. Only without the ants allow revenge.
Now that the model is ready – we can “chat” it – send it a question (fast) and it returns what it calculates to fulfill it (answer). The “deeper” and “subtle” model is – the better the answer. And so – if we want to improve the answers in a particular “hidden” area (“Gray hamster sleeping habits in the eastern part”), we need to feed the model more about this topic – may – less publicly accessible – content. This is called “fine -tuning”.
The amazing thing about this is that the model actually represents human knowledge. At least what is available in writing. This “knowledge” does not equal “truth” and it develops over time. Thus, a similar tactic is used, such as what Google did back with Pagarank a day to improve the reliability of the model return. By giving more weight to reliable and actual sources. We are in the early days of these models – so wait for the subtle tuning to be less a matter (since more “edge” content goes to the foundation models) and we hope to increase accuracy (keep in mind that we used many scaming sites in Google?).
Communication becomes completely rag
So where can it be related to all this?
At first glance, the GPT and other foundation models are “better search engines” and save us time reading, understanding and writing. It will definitely save a lot of time. But is it all?
Hopefully not.
Broadband connection between humans and human knowledge (now I abandon ants), as is represented by foundation models, can (and should) miss out on language restrictions. When it is constantly refreshing, a cloud of glossy, nuanced, layered and complex knowledge – why do we have to reduce communication with narrow, mistakes, slow and outdated communication methods, language, language, language, language, language, language, language?
The first step should be to check the facts
Imagine reading the news article and knowing immediately whether the evidence supports the allegations. If you see the story of a new magic-in-chance-how does this reliable research stand? What about the price you see about the product? Is this the best available?
That was the first step. When knowledge controls you (speech or text) a narrow thing and give you facts about it. The lens of the truth in the world you interact with. For Shazam facts.
The second step is subjectivity
When you interact with the world, you come to the context. Your subjective person, what you care about, your priorities and the people around you will change. This “topic” is a complex array of information, difficulties and relationships. The more “knowledge” is taken into account – the better. New methods like Rag and its brother Arag are in the right direction.
The third step is the agency
Foundation models are working today that they do not work. I mean – they are lazy – they don't work for You. You have to tell them what you want to get something from them. This is also limited by both of what you need to know, what to ask and ask (language). The solution is an agency – an initiative to take a knowledge – based on what it feels (in its widespread sense) and communicating or acting on your behalf.
This part is a little further away on the technology path, but not yet. Today's GPT seems to “know” how to write a not-so-so-called code. The code can work on computers and computers are known to be useful when coded. Lack forward – the knowledge is growing in the agency.
Are we again in science fiction?
Language as the only means of communication has served us well. 40,000 years and the governance of the planet is not bad. But this has shown signs of weaknesses related to modernity challenges. For 300 years, this is obvious to thinkers, but remained a problem that had no practical solution. Many business and political leaders also consider language restrictions, not the problem, but rather the advantage. However, these fragments of humanity may have the opportunity to relax.
On the recent rise of foundations in the nucleus of the text generative AI, the ability to convey the truth, treat subjectivism and create an agency. They require smart implementation, but they are based on current and work. It's not science fiction. The only risk is that these technologies hit the quality of the wall (cannot find its Pagerans), is too expensive to scalize or regulate them. If so, these are only delays. If we thought that recent centuries show rapid development and the state of mankind has never been better – the acceleration of hope is right around the corner.
Closing words to calculators and lazy GPS users
“Doesn't the availability of all knowledge make us stop reading, learning, thinking and inventing?”
I think not.
Pocket calculators (and supercomputers) did not stop math and relying on family doctors did not make us move around. The advancement of the technique will release us from the making of secular tasks. The only thing that changes is what we think of the secular. Crafting words in smart ways to convey a message is something to strive for. But so he cleaned his clothes and preserving food.
On the other hand, initiation, idea, motivation, aspiration, inspiration – they are far from replaced by technology.
This is also the reason why we do not have to be afraid of the rise of a-sediment. As mentioned above – existing technology does not have an agency, not to mention motivation. Or intelligence. It is just the well -organized ants for all words.
