AI 2023. Meet ChatGPT. - page 107

New comment
 
Реter Konow #:
My reasoning is directed at:

(1) Braking the trend of delusion generation by the mythological depths of society's subconscious.

(2) Developing a sober stance towards LLM and image generators (diffusion algorithms).

(3) Assessing the impact of these technologies on society in general and the labour market in particular.

(4) Defining the boundaries of LLM development and implementation.

Progress (whatever it may be) cannot be stopped - it will continue with or without us. we should not slow down anything - it is pointless and unproductive, we need to benefit from technologies and look to the future.

I don't think anyone in their right mind would "animate" modern AI, there is neither soul nor mind there, but AI does not become less useful in everyday tasks. the efficiency and productivity of human intellectual labour is already increasing, it's great. why all these arguments about the "limits of development and implementation" - it's not clear.

 
Ilya Filatov #:

Oh, no. What's going on in the world goes far beyond the sensory capabilities of man. So he's always dealing with a reflection of only a part of what's going on outside. He judges what's going on out there by that reflection. Without ever encountering it directly. Distortions arise along the path of "sensation-perception-consciousness". Moreover, the "picture of reality" is not formed here and now, but is formed from gradually accumulated signals. And if at the same time the world has time to change, changing signals, the understanding of the world always has a lagging character (this is, for example, very actual in cognition of complex things, phenomena and other people). It is like with the image of the sky, the light from the stars of which has been travelling to us for millions-billions of years, we see only the past.

Of course, a person's own sensory capabilities are not enough to cognise the objective world. The world is much richer than we see, hear and touch. Our perception paints a distorted (subjective) picture of reality.

However, in my reply I only emphasised that the world does not exist "in the head" but is "reflected" in it. And you rightly added that it is partially. I agree.

In this context, the logical conclusion is that, unlike humans, the LLM has no direct interaction with the world, it does not reflect it at all. Instead, the LLM "reflects" the human's reflection of the world, which it has previously encapsulated in text.

Double reflection is double distortion, but since much of the emotional subjective "compost" is reflected, in reality the distortion is much greater. So to know the world through the LLM is to distort it beyond recognition?
 
Andrey Dik #:

We cannot stop progress (whatever it may be) - it will continue with or without us. we should not slow down anything - it is pointless and unproductive, we need to benefit from technology and look to the future.

I don't think anyone in their right mind would "animate" modern AI, there is neither soul nor mind there, but AI does not become less useful in everyday tasks. the efficiency and productivity of human intellectual labour is already increasing, it's great. why all these arguments about "limits of development and implementation" - it's not clear.

1. I have no goal to stop technical progress.

2. There is no goal to stop the flow of mythological nonsense in society either, but there is a desire to slow down the nonsense in this thread in order to preserve sanity and get a useful result from interaction with rational halves of interlocutors' brains).

3. Whether or not the efficiency of intellectual labour with the use of LLM is increased or not should be checked by practice. There is no sense to state it verbatim. There are reasons to believe that it does not.

4. Reasoning about the boundaries of development and implementation of technology is aimed at establishing unoccupied niches and promising areas of development. It is also aimed at identifying dead ends.

 
Реter Konow #:
1. I don't have the goal of stopping technological progress.

2. There is no goal to stop the flow of mythological nonsense in society either, but there is a desire to slow down the nonsense in this thread in order to preserve sanity and get a useful result from interaction with rational halves of the brain of interlocutors).

3. Whether or not the efficiency of intellectual labour with the use of LLM is increased or not should be checked by practice. There is no sense to state it verbatim. There are reasons to believe that it does not increase.

4. Reasoning about the boundaries of technology development and implementation is aimed at identifying unoccupied niches and promising areas of development. It is also aimed at identifying dead ends.

1. then these considerations are incomprehensible.

2. who is delusional here? I haven't seen anyone here who would "animate" AI.

3. it increases. i speak for myself. i throw a piece of code and ask you to optimise the code taking into account the requirement to increase speed and get the result. simplification, code acceleration, help with logical constructions in the code, all this saves a lot of time and increases the quality of the result.

4. While you are discussing boundaries and niches, everything will already be occupied.

 
Andrey Dik #:

...

3. increases. I speak for myself. I throw a piece of code and ask you to optimise the code taking into account the requirement to increase speed and get the result. simplification, code acceleration, help with logical constructions in the code, all this saves a lot of time and increases the quality of the result.

...

If your experience with LLM proves to increase the efficiency of intellectual labour, it can be taken into account. Whether this is true for professionals from other fields remains to be seen.

A certain "surge" in performance can be expected in many fields. But, this does not mean a revolution of intellectual labour. It is desirable to examine the issue in detail, and I will try to do so later.
 
Реter Konow #:
If your LLM experience proves to increase the efficiency of intellectual labour, this can be taken into account. Whether this is the case for professionals from other fields remains to be seen.

A certain "surge" of efficiency can be expected in many fields. But, this does not mean a revolution of intellectual labour. It is desirable to examine the issue in detail, and I will try to do so later.

Architects, landscape designers - anywhere and anything can be improved and optimised with the help of AI. I am not talking about a complete replacement of professions, but as an effective tool in work it is quite suitable. And is a revolution necessary?

Take architects or designers - can you imagine what it is like to draw a building, machinery, in detail? Nowadays it is easier and faster to do it, the initial base is created with the help of AI and a person works out the details.


Or, in sci-fi films we see huge cities-ships, I always thought - how much labour and time is needed to work out and create such a giant in reality... and now it is no longer a fantasy, such cities-ships with artificial gravity (due to rotation) can be developed and built in the shortest possible time. and everything that does not touch - everything will be much more achievable and accessible.

 
Vitaliy Kuznetsov #:

Link to original news story - https://www.insider.com/chatgpt-passes-medical-exam-diagnoses-rare-condition-2023-4?amp

And why do you need a language model in diagnosis? The usual programme would cope no worse. Enter symptoms in the field, press "enter", get the result in the form of a list of probable diseases.

An ordinary neural program would do the same job.

Moreover, back in '14, IBM's Watson language model showed similar results in diagnosing diseases. It could also talk and answer questions. Won a TV game of brain teasers. Where is it?
 
Реter Konow #:
Why do you need a language model in diagnostics? A normal programme would do just as well. You enter symptoms in the field, press "enter", and get the result in the form of a list of probable diseases.

A regular neural program would do the same job.

Moreover, back in '14, IBM's Watson language model showed similar results in diagnosing diseases. It could also talk and answer questions. Won a TV game of brain teasers. Where is it?

So? The neuron passed her medical exam. Where's that neuronka? Forgotten, it didn't have mass adoption. And GPT4 (then 5) will soon be in every iron.

That's the difference. AI is now a trend that will be implemented everywhere.

Better ask yourself: if GPT4 and MJ5 will be in everyone's hands, in what areas will they be used?

Good thing doesn't mean popular, popular doesn't mean good. And if it is both, then the world is changing.

 
The funniest thing was that one respected Canadian scientist, the "father" of neural networks (I can't remember the name), after watching the TV show, said that it was a hoax and a neural network can't answer questions like that. Watson then took a lemon of quid.
1...100101102103104105106107108109110111112113114...215
New comment
Reason: