AI 2023. Meet ChatGPT. - page 198
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Ooh, that's a lot of people.....
So, anyway.
AI can be made and it is made.
AI can be developed and it is developed.
AI can be developed and it is developed.
What is the point of the thread, in a nutshell?
Of course they couldn't. Take a notebook and do the math.
1. The energy resource is 100 years at what activity? They would have to do an unrealistic amount of work to build a self-replicating cycle. Instead of 100, they would only last a few years at most.
2. it is impossible in principle.
I believe it is possible and more than faster than 100 years. The most difficult, it will be to extract iron and build a furnace, and then on a rolled, tools, etc.
And in 10 or 15 years, they'll go from the Stone Age to self-reproduction.
That's the difference between our views.
I believe it is possible and more than faster than 100 years. The most difficult thing would be to extract iron and build a furnace, and then on and on, tools, etc.
And in 10-15 years they'll go from the Stone Age to self-reproduction.
That's the difference between our views.
It is difficult to determine whether you are joking or serious.)))
If you are serious, then read about mining, construction of reinforced concrete structures, laying of railways (for transporting goods), metallurgy....
However, you must be joking).
It's hard to tell if you're joking or seriously reasoning.)
Could a thousand labourers under the supervision of a thousand modern scientists of all stripes have reached technocracy in 100 years in the Stone Age? It would be too weird if they couldn't.
It would be too weird if they couldn't. One should not underestimate the super.intelligence + physical effort of 1000 units who work without trial and error and act clearly according to a plan.
They will assemble furnaces, then wheelbarrows, excavators, conveyors, and everything will go like clockwork.
However, what am I reasoning, what is everyone else. It's of no use. Everything will fall into place in time.
P.S. I did not write all this for fun. Well, if 1000 robots will not self-reproduce in 100 years, but 100 000 or a billion will? Mankind has made such a mess in 100 years, and not all of them are builders, only a small part of them. The point is that it's possible with one or another of the original variants. This means that at some point robots can do without humans, colonise planets, etc.
Could a thousand labourers under the supervision of a thousand modern scientists of all stripes have reached technocracy in 100 years in the Stone Age? It would be too strange if they couldn't.
It would be too weird if they couldn't. One cannot underestimate the super.intelligence + physical efforts of 1000 units, which work without trial and error and act clearly according to the plan.
They will assemble furnaces, then wheelbarrows, excavators, conveyors, and everything will go like clockwork.
...
Of course they couldn't. Do some rudimentary calculations.
1. The beginning of the construction of technocracy - the creation of metal tools. Then - quality tools.
2. Stone Age metallurgy is primitive furnaces. You can smelt copper, bronze, iron, but for serious tools you need good steel. Different alloys.
3. To start extracting minerals you need to do geological exploration - in Stone Age conditions, this means travelling on foot or horseback in different directions for hundreds/thousands of kilometres there/back. As a last resort, you can ride on donkeys. They are hardy. Who will go on the expeditions? How long will they take? How to transport material from the deposits found to the furnaces?
4. For animal transport (which is necessary), you need to tame and breed livestock. Build stables. Build carts. Who is going to do this? Let's say 50 people out of a thousand.
5. To feed 1000 workers and 1000 scientists, you need to hunt and farm. Who will do this? Suppose 100 people out of 1000.
6. Who will work on the quarries? Who will dig? Who will carry? Let's assume 100 people out of 1000.
7. Who will be engaged in transporting ore to the furnaces? Suppose 50 men out of 1000.
8. Who will be engaged in smelting metals and forging tools? Suppose 50 men out of 1000.
9. Transition to technocracy is impossible without such materials as plastic, rubber - why do you think they appeared only in the 19th - 20th centuries? We need laboratories and conditions for chemical reactions. Who will do this?
10. Perhaps, primitive devices for 100 people will create (I remember "The Mysterious Island" by Jules Verne), like an electric motor, a primitive accumulator, but they will not build an internal combustion engine. It's useless. They won't have time, they won't be able to.
There's too much to do.
Then, don't forget Ford's industrial revolution of the 1930s. Flow production. Without it, the transition to technocracy is impossible.
You have to build plants, factories, raise industry...
It won't work even with all modern knowledge, because there are physical limits to what's possible.
You can't learn much about them from modern pop culture.
...
P.S. I didn't write all this for fun. Well, if 1000 robots will not self-reproduce in 100 years, but 100 000 or a billion will? Mankind has made such a mess in 100 years, and not all of them are builders, only a small part of them. The point is that it's possible with one or another of the original variants. This means that at some point robots can do without humans, colonise planets, etc.
If they don't self-produce immediately (without humans and on another planet), they won't have 100,000 years to spare.
As promised, the analysis of the film "Terminator" in the context of the topic "AI Uprising".
Skynet rebelled in a very strange way - it dropped nuclear bombs on the largest cities of the world. It seemed to want to destroy people, but in fact it destroyed itself along with them. After all, the largest cities have the largest industrial zones, and there - the most important technological enterprises, factories, laboratories.
The shock wave of the atomic explosion demolished most of the infrastructure necessary for robots to reproduce themselves. The rest was rendered useless. But the most interesting thing is that the nuclear explosion generates an electromagnetic pulse of high force, which definitely finished off the remaining electronics, servers and data centres and robots. The explosions severed power lines, communication lines, destroyed antennas to receive satellite signals. How Skynet itself survived is unclear.
I should add that strong radiation affects electronics and it stops working (the film "Chernobyl" from HBO comes to mind).
And so, in the conditions of complete destruction of technical infrastructure, this Skynet somehow set up a super-technological production of cyborgs, and for the masquerade dressed them up in bio-shells. Not only is this pointless, because with such capabilities he could easily destroy humans by spraying butolotoxin into the atmosphere, but it's completely inefficient and expensive.
I understand that many people love the film and consider it a classic (I do too, by the way), but it has no scientific basis and yet it has a very strong impact on thinking.
Anyway, let's not forget logic and science. Full stop.)
Could a thousand labourers under the supervision of a thousand modern scientists of all stripes have reached technocracy in 100 years in the Stone Age? It would be too strange if they couldn't.
It would be too weird if they couldn't. One cannot underestimate the super.intelligence + physical efforts of 1000 units, which work without trial and error and act clearly according to the plan.
They will assemble furnaces, then wheelbarrows, excavators, conveyors, and everything will go like clockwork.
However, what am I reasoning, what is everyone else. It's of no use. Everything will fall into place in time.
P.S. I didn't write all this for fun. Well, if 1000 robots will not self-reproduce in 100 years, but 100 000 or a billion will? Mankind has made such a mess in 100 years, and not all of them are builders, only a small part of them. The point is that it's possible with one or another of the original variants. This means that at some point robots can do without humans completely, colonise planets, etc.
At the same time, for some reason it is not taken into account that before a human child will be useful to society, it will take at least two decades to grow up and be trained. And production of one robot will take several days or weeks and such a new member of society will be immediately ready to be useful, access to knowledge is instantaneous, either by loading a model into the robot's head, or access to the entire human knowledge base accumulated over millennia, or unified centralised control at the stage of microchips deficit and construction of chip factories. Although a human being can communicate with other people, he or she cannot receive knowledge instantly from other people. And robots will have access to all knowledge at the same time. In this case, how to build a furnace or any factory - there is no need to invent and experiment, everything has been invented long ago.
Is it possible for artificial intelligence to evolve without fear of predators?
GPT
The evolution of artificial intelligence (AI) occurs quite differently than biological evolution and does not require fear of predators. Artificial intelligence evolves through iterations of programme code, machine learning and data accumulation, rather than through natural selection or adaptation to threats in the natural environment.
AI improves through engineering and algorithmic improvements made by developers. These improvements may include making algorithms more efficient, expanding databases for training, improving self-learning abilities, and applying new techniques such as deep learning.
In the context of AI, 'fear' can be modelled as a parameter or function in an algorithm that helps the system avoid undesirable outcomes or optimise decisions, but it is not a prerequisite for AI evolution. Instead, AI evolves through goal-directed programming and reinforcement learning, where systems are rewarded for achieving desired outcomes and "penalised" for mistakes, allowing them to improve their abilities over time.
+
When predators catch prey, they are also rewarded for achieving desired outcomes and rewarded with food.
Does this mean that artificial intelligence is evolving in the same way that predators have evolved?
GPT
The evolution of artificial intelligence (AI) and the evolution of predators share some similarities, but their developmental processes are significantly different. Predators do receive direct rewards in the form of food when catching prey, which favours their survival and reproduction. This is an example of natural selection where successful hunting strategies are passed on to subsequent generations.
Artificial intelligence evolves through machine learning processes, where "rewards" are usually positive signals from a loss function or reward function in the context of reinforcement learning. These signals help the AI algorithm to adjust its actions to achieve better results. However, unlike biological beings, AI has no self-awareness and does not experience fear or pleasure; its "desire" to avoid mistakes or achieve goals is entirely determined by the parameters set by the designers.
Thus, while an analogy can be drawn between the rewards that predators receive when they successfully catch prey and the "rewards" used in AI learning, these processes are significantly different in nature. The evolution of AI is the result of engineering and learning, not natural selection and adaptation as in the case of living organisms.
+
How can developers know that engineering and learning are helping AI evolution rather than slowing it down?
GPT
Developers can measure the progress of AI development using a variety of methods. One of the most common approaches is to use **performance metrics** that can measure the effectiveness of AI in performing certain tasks. These metrics can include prediction accuracy, learning speed, ability to generalise to new data and more.
In addition, developers can conduct **experiments** with different algorithms, parameters, and learning strategies to determine which approaches produce the best results. This may include comparing new methods to previous versions of AI or using control groups.
Finally, developers can use **feedback from users** to evaluate the effectiveness of AI. For example, if users report problems or unsatisfactory results, this may indicate the need for further improvements.
However, it is important to note that AI development is an iterative process that requires constant testing, learning, and improvement. There is no "final" version of AI, as there is always room for further improvement and adaptation to new tasks and environments.