Artificial intelligence will change everything.
I mean the world of artificial intelligence and solutions, which only recently seemed to be a domain of science fiction films.
The biggest, most spectacular impact on the changes was the emergence of a free online GPT 3' chat app.
It turned out that some free, online app is able to work more accurately and effectively than many professionals with years of experience.
This caused a real earthquake in the labour market.
W-MOSZCZYNSKI-2024-4-36The Pandemic changed everything.
The Pandemic has brought a great deal of interest to the work of IT analysts, data analysts, programmers and database administrators, thousands of people trapped in their homes.
People were working in their homes, and there was a wide range of communication, and it turned out that you could organize remote meetings with a camera and a microphone, and today these statements sound cliché.
Until a few years ago, meetings where we saw people on computer monitors were the domain of science fiction movies.
The Pandemic broke the social taboo, showed how important communication and scattered cloud systems are.
The advent of intelligent language models
The first symptom of the changes was a little unechoed new algorithms of Google translator.
Before 2016, the texts translated by the Google app were characterized by low quality, lack of flexibility and general meiseoty.
A similar language interpreter integrated with Google translator worked. Translated sentences were very cognizable, the application only translated individual words.
She did not understand the context, the ambiguity of words. In other words, this algorithm was not suitable for professional use before 2016.
Everything changed the introduction of a new Google translator algorithm based on a smart language model of a large scale®.
Currently, Google translator is able to translate the spoken language directly and transform it into an excellent translation quality.
Language skills of this device are the subject of research and analysis, scientists by
and Currently the newer version of this chat GPT 4 is common
2. https://translate.google.com/
3 Large-Scale Language Models (LLM) are a kind of artificial intelligence (Al), which uses deep learning techniques and huge data collections to understand,
Summing up, generating and predicting new content. These models are trained on huge amounts of text, which allows them to learn human language in a statistical manner.
for many years they could not believe in the perfect results of this application.
Large-scale language models were the basis for the introduction of intelligent language models, including the popular GPT chat.
What is a GPT chat?
Chat GPT is a large-scale language model, which, like most of these models, is based on the Markowa algorithm.
It's a simple algorithm that the appearance of one word implies the probability of another word being present within words.
It's a mechanism that's been expanded into a model for great operational capabilities.
The widespread use of ChatGPT has resulted in significant reshufflings in the labour market, not only in the IT sector but also in the labour market of graphic designers, as is widely believed,
designers or creators of texts.
What can ChatGPT do?
Access to the free model of the GPT 4 chat has virtually everyone. Just install your Bling browser on your phone.
You can enter any wish in the window, which will be fulfilled very quickly.
So you can ask the GPT 4 chat to write an article on food market problems in Poland over the last 7 months.
You can mention adding appropriate links
4 In 2018, a group of researchers from the University of Tokyo published an article,
in which they compared the accuracy of Japanese poetry translation by Google Translator with translations made by people.
They found that Google Ttumacz was able to correctly translate the basic meaning of poems, but he often made mistakes in translating more subtle language nuances.
Information generated by the AL gemini.google.com model.
5 Markov's algorithm was first formulated by Russian mathematician Andrei Markov in 1907.
Markow published his research in his work entitled
Markov's algorithm is a stochastic model that describes a sequence of events where the probability of each event depends only on the previous condition.
In other words, the future is conditioned only by the present, and the past has no bearing on what will happen in the future.
Definition generated by the model AL gemini.google.com.
and sources for specific press or television publications.
At the same time, it is possible to define the form of this article, whether we want to receive a serious scientific text or a joking form of publication full of anecdotes and fun examples.
ChatGPT 4 is able to program and write programs in different languages. It can be given a specific optimization task and directed to specific databases.
Based on the data description, the GPT chat will programme the entire model in Python, present the results of the calculations and comment on them.
In word, he will perform the work of an experienced developer and data analyst.
If we have written the code ourselves, the GPT chat can check the correctness of the code components, it can also find errors that have slipped into our application.
This is not the end of the possibility of a free application installed on our phone. Chat GPT can also create interesting graphics or can make descriptions for products.
Let's say we have 700 different foods in our warehouse, and we want to create an online store to sell our goods online.
It is not difficult to paste pictures into the sales site. It is harder to describe each of these products. It also can come with the help of the GPT chat.
Just show him pictures of individual products, and he will automatically create a description of them.
We can also influence the content model, in points we can write what should be in the description.
So we can type in terms and numbers like diet bread, 12% glucose content, 7% fiber content, some other product characteristics.
Chat GPT 4 based on photos and words will create a coherent description.
Application of ChatGPT 4 in employee recruitment
The direct effect of the widespread use of large-scale intelligent language models (LLMs) is to reduce the employment of IT professionals, graphic designers and other professions,
which have been or are about to be automated by free GPT chat applications.
An example is the recruitment industry.
Before the emergence of intelligent language models, the recruiter had to analyze dozens and even hundreds of CVs that were sent to a specific workstation.
His work consisted in searching for the appropriate words that fit the specific requirements contained in the advertisement for a recruited post.
Today, the same work will be done by the GPT chat; just give him the resumes, and he himself will create a summary or select 10 candidates,
who best meet the conditions set out in the notice.
This example implies two simple conclusions:
1. You can fire the person who verified your resumes because the GPT chat does it faster and better and above all cheaper.
2. Those who apply to specific workstations must edit their CVs so that they contain keywords (so-called meta tags).
The automatic language model will analyze the density of keywords and how they are used. In other words,
The effectiveness of our application for a specific job will depend on the quality of the CV created. The quality of the CV will depend largely on the ability to use keywords.
You will need to use techniques similar to web positioning. Who will best position meta tags? Of course, the language model AL.
After some time, CV verification models will detect and eliminate enhanced CVs, just as Google's browser algorithm does today.
So the models that create the CV will avoid these security features, and it's going to create a vicious circle of artificial intelligence models.
ChatGPT and the invention of automatic looms
A good example of the phenomenon of drastic social changes caused by technological changes described here is the example of the implementation of automatic looms in the 19th century.
Used to weave carpets, patterns, decorated shirts were used by highly qualified specialists.
Automatic wood could weave the most complicated patterns with enormous precision and it did so at a turning rate.
From day to day most well-paid weavers, specialists with great practice and knowledge lost their jobs. The prices of fabrics and stitched materials have fallen dramatically.
New competitions of automatic loom programmers, mechanics have also emerged.
Who's gonna check the financial reports?
Another example of changes is the area of financial report control.
In the past, financial reports could have been verified by outstanding specialists in financial law, economics and finance.
Analysis of reports is essential for investment banks, tax advisors and equity funds. This is an important element of modern market economy.
Now just paste the financial report into the GPT chat window and write what our expectations are, and artificial intelligence will generate any summaries, find errors,
it will express its doubts and even enrich its report with charts and prediction models. Something that still took several days ago now takes several seconds.
Tutoring with the machine?
Intelligent language models can be found in the most unexpected places, industries and sectors. For example, in the tutoring industry, in conversations in foreign languages.
Once it seemed that nothing could replace learning a foreign language by talking to a living person,
The automatic, self-constructed weaving of French Josekrosno was built in 1890.
6 The automatic weaver looms, which allowed for the even and multicolored fabrics, became Pha M. Jacquard in 1805.
Natom in which the thread is during weaving the American mechanic James H. by model AL chat GPT4.
The most important is the lack of clear feedback.
If we speak English to a British citizen, we can make a lot of mistakes, have a bad accent, and use wrong words.
And we won't know exactly what our English level is.
We're gonna fix the mistakes, still speaking with the wrong accent and using strange sentences.
It turns out that one can talk freely with ChatGPT.
You can preprogram the pace of conversation, the range of vocabulary used, we can finally indicate if the GPT chat can interrupt us and correct our mistakes.
Most importantly, all the content that we will say in English and the participation of the GPT chat in our conversation in English will be recorded in the form of a register.
Each sentence, each word used will be written down.
We can ask the chat to emphasize all errors or to write a more correct form of sentences in parentheses next to our sentences.
This means an increase in the quality of conversation in English or in any other language. Chat GPT always has time for us, is patient and costs nothing.
After each such conversation, we can independently verify the quality of our language skills by analysing the conversation log and the AL model can also be analysed for us.
What are we supposed to do when high-level models are spreading?
As with the mass introduction of automatic looms,
also in case of the spread of large-scale intelligent language models (LLMs) there will be mass dismissal of professionals of most professions.
So far, I have mentioned that mid-level specialists, such as programmers, builders of content or graphic designers, can lose their jobs.
Unfortunately, mass redundancies can also affect high-level specialists such as doctors, food technicians or tax advisors.
The threat is the emergence of language models of the GPT chat type combined with so-called expert models.
The idea of establishing expert models emerged in the second half of the 20th century.
The expert model in the simplest version is that further questions are asked and answers to them lead to a set of further questions from a specific process.
For example, an expert model may be a virtual doctor who asks the question: Does your throat hurt?
Answer: "Yes, it leads to further questions: Do you have a cold, or do you have a higher temperature?"
Another important combination of the expert system and artificial intelligence will be the introduction of integrated accounting systems that will make their own accounts,
reconciliation of balances, creation of financial statements.
Specialist professions that seemed to be safe,
Now they may be completely unnecessary because they will be replaced by advanced models of artificial intelligence.
A machine better than a human?
We can always be comforted that the machine will never be more intelligent than man. Unfortunately, now machines are more intelligent than humans.
Of course, in terms of specific specialties.
A good example is the identification of photos. Models of photo identification based on neural networks can identify the most illegible and ambiguous photos.
They can assign plants or food products to specific species, their origins and even indicate hazards,
such as unseen human symptoms of mold or expired.
Food companies can, of course, continue to maintain specialist positions, such as food technology.
However, if the expert systems combined with artificial intelligence are to perform the work of food technologies in a more precise and cheaper manner,
technologists will no longer be needed, even if some humane company doesn't fire the techno, it will do the competition.
If competition has lower production costs due to lower labour costs, it will achieve a competitive advantage. This advantage will lead to the collapse of the company's humanitarian.
The economy is a connected vessel system.
This is just the beginning.
Development and mutual competition between models of artificial intelligence is just beginning. Currently, several competing intelligent language models can be mentioned.
Model manufacturers announce further versions of their products. They are brilliantly talking about the fact that the next version of the model will be many times better than the previous one.
Just compare the quality of the response that was run by the Chat GPT 3 version with the quality of Chat GPT 4.
We know that intelligent models have existed in economic space for only two years, so it can be said that this technology is in its infancy.
At the same time, it can be said that intelligent models perfectly replace people's work.
After the introduction of an automatic loom, one person was able to perform a daily job comparable to that of 30 qualified people.
Intelligent models can multiply one person's work in a similar way, and that's just the beginning.
The perfection of these models will grow in the exponential system. The second model will be twice as good as the previous one.
'The third generation of the model will be 5 times better than the second level model. Why is this happening?
Because creating another version of the model is done using an earlier version. In other words, models form themselves.
Twice the perfect model can create a four-fold better version. Four-fold the better version I will build a 10-fold better generation of the model.
Man is no longer needed. Today, the creation of a model is fully controlled by man.
The next generation of higher models will be created by subsequent models, man will understand less and less and have a smaller share in the technology of creating new models.
What's the specialist's role now?
As specialists, we should not stand up for the domination of the human species, but we should not sabotage rapid technological development, but we should encourage it.
Everyone is able to participate in the technological revolution, just find an area that can be multiplied in terms of efficiency using smart models.
Access to smart models is very easy. Today, many people use web browsers.
Web browsers usually create user identities. The most common type of identity is Google identity.
You have at your disposal an email with the extension of gmail.com, and you also have Google disk in the cloud.
On a virtual disk it can create and work in Google spreadsheets, use Google text editor, create presentation and other necessary things.
Also Google has its smart model, as good as the GPT 4 chat. Just log in to the site
gemini.google.com. After logging in, a window will appear in which we can enter any text, query or give some document to check.
It may turn out that our laboriously performed duties can be done very quickly and with much better quality.
The application is free, available on virtually every device. We can experiment, we can use guides that can be found on the Internet.
It is worth taking advantage of this opportunity to develop a new, self-specialization. Undoubtedly, intelligent models will cause large perturbations in the labour market.
It is worth being on this side, which will benefit from the benefits of the new technology.
Wojciech Moszczyński — graduate of the Department of Econometrics and Statistics of Nicolaus Copernicus University in Toruń; specialist in econometrics, finance, data science, and management accounting. He specializes in the optimization of production and logistics processes. He conducts research in the area of the development and application of artificial intelligence. For years he has been engaged in the popularization of machine learning and data science in business environments.
Dodaj komentarz