What happened to the AI revolution?

Several, even among others The Economist have started to ask questions, whether and where the artificial intelligence revolution has reached today? The issue was also discussed yesterday in Mektory From the AI/ML forum. The billions of dollars, euros and other units invested in all kinds of language and other AI models have, according to many, not yet produced any economic impact, etc.

But the fact that the field of artificial intelligence has already spawned countless AI experts has even reached the world of memes.  

After the forum, where panels and discussion groups discussed the impact of AI in medicine, education, media and elsewhere, I will also share some of my thoughts.

AI experts or enthusiasts?

I don't consider myself an artificial fur expert, and with all due respect, I don't know anyone who is. This is primarily because AI is a very broad field and it is not possible to know a lot about all topics. At the same time, I think that if someone has tested a solution for a long time and thoroughly and has also benefited from it, then he can definitely consider himself an expert in, for example, Chat GPT or some other AI tool.

A revolution in less than two years?

Chat GPT was launched less than two years ago on November 30, 2022. Its first versions did not attract the masses. In reality, the biggest steps of generative or creative AI solutions can be seen only from last summer. Although the development has been insane, there are still many today who expect more from the moment. 

I got my first phone in 1997. It took more than ten years before other functions began to appear on the phone, which today almost everyone uses more time than the original function of the phone, i.e. calling. Many websites get more clicks from phones than from computers. Moreover, the development of websites can be viewed in the perspectives of more than ten years. 

AI images have made more progress in two years than photography has in 200 years. Very great results have been achieved in fields that require large amounts of data, such as weather forecasting or medicine. For example, a week ago it was announced that thanks to the artificial branch, it is now possible to detect cancer at the nano level. 

The "Super Brain 1" or "seventh sense" made by Estonians and intended for the visually and hearing impaired, which I was also able to try, is a solution to which AI has made a significant contribution and has great potential to improve the world. 

7sense.ee SuperBrain 1

The Gartner Curve and the Future of AI

"Gartner Curve" is a graph that has also been used a lot in the field of artificial intelligence to explain the current situation. I made a simplified version of it where the point remains the same. 

We will probably also see a situation in the case of AI, where high and fast expectations will first lead us to the "valley of disappointments", but in the coming years we will be enlightened, i.e. artificial intelligence solutions will consciously become part of the daily life of the larger masses. Be it AI agents that help learn, make plans, automate tasks or other solutions. 

AI/ML Forum – so what was said anyway?

Several conferences with presentations, discussions, etc. are nice and exciting, but a few days later you may not remember anything special about it. In fact, it can be difficult to take part in everything, for example in Mektori there were panels on the main stage all day long, but in addition to them there were also round table discussions on a total of 14 different topics in two or three rooms at the same time. 

It wasn't until I had moved on from the workshop on AI education topics to the use of artificial intelligence in the media that it occurred to me that I could actually have started AudioPen or the phone's audio recorder to get a transcription from it, which later could be used to make various summaries using ChatGPT, Claude.ai, etc. - "name a way the main point discussed in the workshop", "make a short summary in a journalistic style, add a title, etc".

Fortunately, the organizers took this task upon themselves, as all the round table conversations were audio recorded as well as a video recording from the main stage.

So, you are at school or at a conference - pick up the phone, record the sound, use some solution (in Estonian, for example for text.ee) and get a transcription of it. You can make quick summaries of this with various AI solutions and find out what was said anyway?

AI in Education

The fact that artificial intelligence is already changing education today is more clear to students than to teachers. However, if teachers could understand that their most deadly job, i.e. correcting students' homework, could be significantly accelerated with the help of AI, tens of thousands of overtime hours would not be done this year, in addition to the whole of Estonia.

AI in media

When is an article written by AI? Can you even say that? If a journalist asks for ideas to write about, do you put something together yourself, but let some language model go over and supplement it? When a photographer takes a picture but alters it with some artificial photo editing? Can some AI footage be added to a filmed clip? Can artificial intelligence be trusted at all for facts?

Try for example chatter.hume.ai – an interactive podcast that you can freely jam into. Think if a similar solution could already read Estonian news and make summaries of them, which it would read aloud. Technically, it is already not difficult to do this in real time in the form of a video, for example with your own avatar. 

All in all, AI is here to stay. Just like the internet or telephone. But its development is faster than any other human invention. Today, it is important to test and try in order to reach Gartner's “productivity plateau” before others in the future. 

I can definitely do something practical and/or fun for your company AI training to offer. Get in touch.