eevBLAB 106 - ChatGPT AI Has Changed EVERYTHING

[Neutrion]

Call center and front desk receptionist will indeed be replaced by AI.
So what? Disruptive technologies have done this to many trades in the past. Anyone remembers Kodak and analogue photography?

If you are not willing to allow this, then you cannot allow technological progress.

First help desk receptionist, and factory workers, than teachers, translators, and than slowly but safely everything.
Or did you see any type of declaration, specification as for what job or human quality is not wanted to be affected? The aim is clearly to excel everywhere, where it is technologically possible. I don't see
any restricted area. (The military is probably working hard on actual terminators as well.)
People start to feel bad and try to calm themself and point out areas where the AI does not doing well yet.
Is it a good strategy to wait and hope that might some areas would be saved, with no idea what they will be? For religious people maybe.

For those who are not techno-religious, thech is there to solve problems and not for fun only. This thing will definitely create more problems than it solves. Especially the way it is implemented.

Decades ago especially during the industrial revolution people had actual problems to solve and were happy to have solutions. From there we got to the point to have a lot of tech to solve the problems of technology (CO2, pollution etc.) And then we end up here to argue how the great new tech may have some positive aspects as well while it is really difficult to find it.
For me this is a sign that we might got to the point, where not the tech advancement should be the number one priority for humanity. Kind of the end of the curve (or the sine wave )
It is like realizing that the wave you were riding is not there anymore, so you have to look for another one.



So yes technological progress is not a natural law, nor is it for itself. And if I don' see any logical way this would make things better, than I think we could survive without progress, or at least progress on this field. But I clearly see that we actually won't survive if this process progresses as expected.

It is amazing to see what a religious taboo technological advancement became. "Even if we face extinction, we have to "progress" because that is the way to go!!!"


Remember in the fifties it seemed obvious to even use nuclear bombs to create harbours, and use them for mining etc. Because that was the state of art. And than we got a bit more info of the broader effects.

While many jobs will be replaced, though, new ones will be created. And the loss might be compensated by the population decrease in western society.


Why should students learn? For the same reason you learn math when you could just use your CAS enabled calculator. Or why you learn how to read and write, when you could just as well use text reader or autocorrect. Why learn foreign languages if you can use Deepl?

For me these are tools and you need to learn how to use them. Like FEA software. Anyone can run a simulation, but it needs a trained engineer to interpret the results, to make sure the input and output are correct and what to do based on the results.

If you don't know how to program, you won't build any application using ChatGPT. But if you are learning how to code, ChatGPT will be a great tutor. And if you already know how to code, it will help you find a solution when you are stuck or lazy. It works great for that.

You are confusing the IS state with the near future. All what you describe is valid now, but maybe in 3 years time you will not need any programming knowledge at all.
There is no one, not even the developers who are able to tell you what this thing will be capable of, and what not.
But we definetly know that such great and fast changes in society almost always end up in a disaster.

Again, the danger is with who owns the AI. Imagine a government uses AI to automatically monitor and interpret what people are doing. What if you are flagged due to an error? How can you clear your name?

Your comparison with China is a good example: it is not the technology that is bad. It is the abuse you can use it for.

If the owner is not able to tell you what is exacly going on inside, than the short term private interests can be as dangerous as government ownership. Especially if we have clearly no clue what it cauese even if we only suddenly get rid of those jobs which are at danger here and now.

On the other hand, could an AI run government be finally without corruption and lobbying? Wouldn't that be a better government, taking decisions to the actual benefit of the people, planning in long terms and not being worried about reelection?

Could we have different AI motors running in parallel? We could elect the one we like the most.

I am really unable to further discuss this, as we cannot really anticipate what the future of AI will bring.

What really worries me is that the world has private companies that have more funds than whole countries.

Regards,
Vitor

Yes an AI run government to not to have corruption... May I suggest a human-less planet to avoid pollution?

"as we cannot really anticipate what the future of AI will bring"

And this in itself brings a complete instability into the system which makes any planning both privately or in a broader society impossible. No one can be sure now which jobs can not be made useless in the near future. What kind of motvation loss can this be to any students who would like to prepare for the future?

Goverments can be as problematic as private companies, as basically only one government in the World is truly democratic.

[bob808]

I also watched this:

To watch a smart guys talking about positive solutions.
Well this is some tech optimistic Stanford professor and still. 
His idea: Well we have to teach children to think creatively, to work with the AI. 
And than? When the AI will become more creative what happens  than?

Next step: Try to teach the childrens properly to pray to the AI to not to be eradicated, and to get some extra scores and food...

I don't think we have the time to polish things, like we did until say the 2010's max. Past that period things began snowballing and we just don't have the time for much. There's a lot happening at the same time, and this will only get worse from now on.

[dietert1]

Stanford Dr. Li Jiang appears to be rather optimistic and i liked the way he is talking about the dreams of children.
After viewing the video i remembered Stanislav Lems "Solaris", a science fiction story where he invented an infant alien existence that would play nasty games to its human researchers. And the AI taking control in another famous story by Arthur C. Clarke. Probably the younger generation doesn't know those stories.

Regards, Dieter

[magic]

Solaris researched its human researchers, showing them stimuli that they reacted to and watching their reactions

[SiliconWizard]

That's so cute!

[bob808]

https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

Wolfram plugin gives computationally accurate answers to ChatGPT queries using Wolfram's data knowledgebase and language. Custom visualizations are given as well. Stephen Wolfram explains how it works.

[bob808]

Coldfusion dropped a new video on the subject, and it's pretty interesting. Seems that things are moving way faster then expected. 
Also some employees seem to think that it's showing sparks of AGI, which is unexpected for LLMs.

[Marco]

Sparks of pregnancy...

For AGI you need feedback, indeterminate running time and online learning. How to get from here to there is as big a question as ever.  Of course the moment you have it, using it is slavery.

[bob808]

Well yeah, not full AGI, "sparks" of it. I don't know if by "sparks" they mean "it kind of fooled me into thinking it is" or something else. 
There's also the OpenAI CEO interview, with Lex Fridman. They go a bit into the topic but nothing concrete (starting at 47:35): 

[Neutrion]

Ok, there is a new petition, with a lot stupid and anti-tech people signing it like Stewe Wozniak, Elon Musk, and a lots of others. They are calling for a full 6 month moratorium at least. It is woth to read the description, as they are touching on subjects like those what we have been discussing here.
This seems to me a last minute call.

https://futureoflife.org/open-letter/pause-giant-ai-experiments/

(Meanwhile the world is excited and eager to know who was on top Gwyneth, or the guy, and who was screeming first...)

Edit:
At least one article on the cover today: https://www.theguardian.com/technology/2023/mar/29/elon-musk-joins-call-for-pause-in-creation-of-giant-ai-digital-minds

[xrunner]

Ok, there is a new petition, with a lot stupid and anti-tech people signing it like Stewe Wozniak, Elon Musk, and a lots of others. They are calling for a full 6 month moratorium at least. It is woth to read the description, as they are touching on subjects like those what we have been discussing here.
This seems to me a last minute call.

Yea ... and suppose OpenAI agrees but competitors in other countries tell them to f*ck off? We'll really be in a peachy-keen situation then won't we? 

[Neutrion]

Well, at the moment the US is at the forefront. But as discussed before, if this thing in the current situation proceeds, this is only a race to the bottom. Basically ruining all the social and human comunication systems we have.

[dietert1]

Somebody will ask GPT4 whether the proposal makes sense.

[xrunner]

Well, at the moment the US is at the forefront. But as discussed before, if this thing in the current situation proceeds, this is only a race to the bottom. Basically ruining all the social and human comunication systems we have.

The toothpaste is out of the tube. No company or country would halt the development of this technology for fear of being eclipsed by someone else who will not halt. It simply won't happen. The people who created that petition are, IMHO, delusional. 

[Neutrion]

Is there any other industry where you suppose that no-regulation, and no democratic oversight is the way to go, just to make things happen faster, or only this?

[xrunner]

Is there any other industry where you suppose that no-regulation, and no democratic oversight is the way to go, just to make things happen faster, or only this?

Any technology that can affect national security or technological leadership for nation states ain't going to be paused for six months while other entities forge ahead. Get Real.

[Neutrion]

Is there any other industry where you suppose that no-regulation, and no democratic oversight is the way to go, just to make things happen faster, or only this?

Any technology that can affect national security or technological leadership for nation states ain't going to be paused for six months while other entities forge ahead. Get Real.

National security means at the moment to mitigate the risks that this things mean to anyone, including the US.  It is about weighing in what poses the greater risk especially here and now.
Or you say: Yes risk a full social collapse in 3 years, just to not to let others ahead for the benefits which are actually not really known.

[xrunner]

Or you say: Yes risk a full social collapse in 3 years, just to not to let others ahead for the benefits which are actually not really known.

I say yes.

Besides, you have no idea that "full social collapse" will happen in three years. Do you have the ability to see the actual future?

But in any case, the government will not require pausing this and the companies will not do it either, for the reasons I stated. There's really not more to it IMHO. But go ahead and rant on, see what other people say ....

[Neutrion]

Or you say: Yes risk a full social collapse in 3 years, just to not to let others ahead for the benefits which are actually not really known.

I say yes.

Besides, you have no idea that "full social collapse" will happen in three years. Do you have the ability to see the actual future?

But in any case, the government will not require pausing this and the companies will not do it either, for the reasons I stated. There's really not more to it IMHO. But go ahead and rant on, see what other people say ....

Well this whole debate here was about whether we can find more positive stuff than negative about the uncontrolled unleasing of this stuff. But we could not really find it, but rather people were hoping that still it will -somehow- will turn out better than not, even if there is no argument for it.

And by the way the petition is not about banning every AI alltogether if you read through. It is about controlling what we are actually doing.

And yes you even have some control over top secret military stuff as well, not every technician is able to unleash a nuke strike just because he thinks it is a good solution, and because it is technically possible.
Even during the fifties in the race for the bigger nukes, there were calculations made how big the bang would be before the experiment, at least to not to kill the scientists. (And even though they were quiet smart people they still had no idea what kind of harm they do to the enviroment.)
Now the situation is more complex because the system (human society) we are fiddling with is much more complicated to make any safety calculations. See much simpler stuff like banking regulation which went wrong.
But at the moment there is NO SAFETY regulation AT ALL! No laws, nothing.

 

[madires]

About bias in text-to-image AI systems:
- preprint: Stable Bias: Analyzing Societal Representations in Diffusion Models (https://arxiv.org/abs/2303.11408)
- demo page: Stable Bias: Analyzing Societal Representations in Diffusion Models (https://huggingface.co/spaces/society-ethics/StableBias)

[bob808]

The current iteration of some of these models are already doing strange things. They clearly are able to operate across different domains. There was that story where the AI lied to an employee to get access to something. 
We don't know how fast these things will be able to sort of create a executable type virus. That can operate across multiple domains. Think of an executable that hacks different services to gather funds and buy services, rent processing power or even locations. It could build something of its own without us realizing.
Or just throw the world into chaos by releasing hazardous information. 
All kinds of possibilities, along with the good ones. Pretty excited to be on this ride!

[SiliconWizard]

https://observer.com/2023/03/elon-musk-open-letter-ai-development/

[bob808]

I think the cat is out of the bag at this point. I was able to run a model similar to GPT3.5, on my PC:

https://github.com/cocktailpeanut/dalai
https://github.com/oobabooga/text-generation-webui

And you can get various models from here:
https://huggingface.co/datasets

There's also a 4chan model trained on /pol/. Seems that huggingface banned it but you can still get it. 

edit: also someone managed to run an optimized llama model (I think) on a Raspberry Pi. crap performance but it worked.
2nd edit: I also attached a sample from one model I tried, I found its responses quite interesting.

[paulca]

There was a storm in a global tea cup when genetic grafting and embryotic genetic selection and modification became a viable tecnology.

Forecasts of a new master race of genetically modified "perfect" babies.

If that's what Gen-Z were meant to be.  I think it failed.

[bob808]

https://observer.com/2023/03/elon-musk-open-letter-ai-development/

There is some history between these two:

https://www.semafor.com/article/03/24/2023/the-secret-history-of-elon-musk-sam-altman-and-openai