Bioenergetic AI Labs: The future of Ray Peat's legacy is already set

[sharko]

@yerrag said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

@sharko

Let me get this straight though.

You say AI lies.

I see lying as a conscious effort to mislead and to deceive. When it gives a wrong answer because the answer is based on being fed wrong data, does that constitute an act of lying?

If you fed it good data, the likelihood it gives you a correct answer is increased, doesn't it? Yet it may still give you a wrong answer, for various reasons. You are right in not trusting it, but will you not trust it because it consciously lies?

When you have two identical instances of the AI mind, will you get two different answers if it is telling the truth? For it to be lying, won't it give two different answers when it consciously lies?

Regardless of the information it is fed or the information it is trained on, all existing tools have a tendency to "satisfy you".

Let's say you have an AI with access to all the knowledge that Ray Peat has published and you formulate a question to it:

"Explain to me the effects of calcitonin on calcium regulation and its relationship with PTH and add 5 quotes from Ray Peat on calcitonin"

He's more likely to come up with 5 citations even though there aren't that many Ray Peat citations on calcitonin, than he is to tell you he only found 1 citation.

And if you write to him in response that he made up the quotes for you and ask him not to make anything up, and if he has no information on the subject, that he write the truth.

There's a good chance he'll try to set you up even one more time.

Generally, the pattern I've noticed in all the models I've tested is that after the second request they level off and stop inventing. But it is not an iron rule.

The more "smart" the model, the more difficult it is to control it with system prompts. In fact, he is more out of control. For example, peatbot according to my test gives very correct answers because it is based on a very basic model and is also fixed only on the knowledge that was added to it according to what I saw. It has advantages and disadvantages. But when you learn to deal with the more "smart" models, their usefulness is much greater

[yerrag]

it u @sharko

I guess I was expecting AI to be programmed like Robocop. If you gave it a directive not to lie, it will follow.

But maybe if you don't force it to lie (or make things up) by asking it to name 5 citations when there really isn't that many, it won't have to be forced into lying. Couldn't you just ask it to give citations without requiring how many, then it wouldn't have e to "lie."

I guess what I'm trying to say is that you should be in control of the behavior of the AI, but what you are saying is that you'd have to lose that control to have a smarter AI. zero I would then prefer a dumber more honest AI then.

[sharko]

@yerrag I understand your preference for a dumber AI.

It is difficult to determine what is best for everyone at this stage but everyone can know what is best for them, only after learning to use it correctly with a "smarter" AI.

Like you said, it really depends on the request. As long as you ask him for information that is available to him at that moment, it is unlikely that he will invent something that does not exist, but if you push him to give you an answer that he does not have, he will prefer to invent and not say that he does not have. This is something that I and millions of other people try to solve every day by "programming" prompts at the system level. You can get nice results, but a week later they can change. Because this is AI and not rigid settings

[sharko]

@ThinPicking said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

@sharko said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

In context, this is a work of art.

Yeah, it's a beautiful piece of bullshit

But honestly, soon it will reach a precision capability that will allow us to see on a visual level anything we ask and it will be crazy

[sharko]

"Create a illustration image of mitochondrial uncoupling"

OMG it looks like oranges in there now everything is clear.

[ThinPicking]

It was the inorgnic phoiohat all along!

[DavidPS]

WHAT HAS THAT GOT TO DO WITH THE PRICE OF TEA IN CHINA?
.
User: what is the price of tea in china?

Peatbot.com: I don't know.

A truthful response, but admittedly not very creative.

[yerrag]

@Kvirion said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

@yerrag said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

Thanks for sharing these ideas. Did you learn these concepts by self-study? Or did these come from a liberal arts education that are similar to Peat's that give such a perspective?

A mix of... partial education in liberal arts plus self-study thanks to some wise people sharing their insights online/books... Plus recently reading/learning from Ray's works (I miss him so much...) helped me to add another dimension... but I'm still learning and I'm open to dialogue

BTW I also received professional training in advanced sense-making and Complex Adaptive Systems.

Can you be more constructive rather than talk down then as if you want to give a lecture more than help Sharko make his efforts of using AI to help us gain a better understanding of Ray Peat's work?

Yeah, you're right, my bad. I may try... but both sides need to be more open-minded...
Honestly, I'm pretty frustrated with the IT guys claiming unfoundedly that they have a panacea or they are helping the world... When in reality they are unaware (WEF/neoliberal?) agents of destruction/idiocracy (i.e. Moloch)...

In my imagination I'm with John Connor, Morpheus, and Butlerian Jihad - fighting the machines!

BTW I tried to be nice to Sharko at first...

Full disclosure - in the ancient past I also worked in roles of IT database developer/Analyst or IT project manager and I was a technology fanboy.

But I may be wrong in not giving you and T-3 enough credit as really I can see some effort to be constructive, and that Sharko's responses to you may not hit the right notes with you, though I get the sense that he is more about explaining the possibilities and potentials if AI, than in addressing the points you raised.

Right, it's good to use/explore possibilities and potentials, BUT one also must be aware of (many) limitations...

LLM can help us find something, but such info must not be seen as a conclusion, but only as an input for further conscious processing with the help of the scientific method and creativity...

But let's continue the diacussion giving Sharko a chance to answer Kvirion's points, and if Kvirion's points are not sufficiently addressed in the current prototypical stage, then we have to consider the likelihood of improving the AI model instead of prejudice it based on previous attempts of AI thst failed.

Golden advice, I'm for it.

Your liberal arts training and your capacity to keep learning is very Peaty. Thanks for sharing your insights that comes out of a critical appraisal of what adds and what detracts from our discussion on a budding field that comes with it mystery and and a fear from not being able to understand it fully for its scope which is hard to limit and compartmentalize.

I have for now only my own thought processes to guide me in my own expectations from AI. I can only think of it as extremely useful as uncovering concealed dots which I have overlooked, leaving me to spend more time connecting the dots in search of solutions where cause and effect is brought back from exile in the current orthodoxy of academia and science.

[sharko]

@yerrag said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

I have for now only my own thought processes to guide me in my own expectations from AI. I can only think of it as extremely useful as uncovering concealed dots which I have overlooked, leaving me to spend more time connecting the dots in search of solutions where cause and effect is brought back from exile in the current orthodoxy of academia and science.

You summed it up perfectly. That's the whole point

[LucH]

@yerrag said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

I can only think of it as extremely useful as uncovering concealed dots which I have overlooked, leaving me to spend more time connecting the dots in search of solutions where cause and effect is brought back from exile in the current orthodoxy of academia and science.

I agree too, towards this analyze.
Yes, you’ve got it: a human being for connecting the dots, the dots that have been selected by IA, thus sparing time and mind tension + energy. That’s already a fine outcome.
But we are not guaranteed against selective misguided tendencies: the tendency to take for granted what is mainstream. The tendency to search for responses among the approved facts by “authorities”.
So, keep in mind IA would be only a useful tool to make the progress easier. Has to be confirmed by another source, as usual.

[Peatful]

Here is audio of me speaking to Ray

It is short
It should be queued up and ready to go

Listen my question,
and his answer

If you miss his point
Relisten

https://bioenergetic.life/clips/433f2?t=3809&c=72

[ThinPicking]

With respect and although the subject is hard, you have the phonology of a whole hearted songbird.

More importantly. Raymond is right. I'm part-time nerdy, locally executed token prediction is cool (and very crude). "Cloud" computing is uncool (and very crude). But Raymond is right.

Even as an adult.

[ThinPicking]

Mr @sharko, what do you think. Why are we in a rush?

[Peatful]

@ThinPicking said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

With respect and although the subject is hard, you have the phonology of a whole hearted songbird.

More importantly. Raymond is right. I'm part-time nerdy, locally executed token prediction is cool (and very crude). "Cloud" computing is uncool (and very crude). But Raymond is right.

Even as an adult.

No no

Im a despised bitter Karen!

Get it right….

————————————————————-

Him offering me or reflecting back to me what I already knew, was so comforting to me during this time

I was so beaten down as a parent

All those educators in our local schools should be ashamed of themselves for what they did these kids

I will continue to hold on to his words of resist
and not give strength to the irrational and nonhuman system

I miss him

[yerrag]

@ThinPicking said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

Mr @sharko, what do you think. Why are we in a rush?

We're moving too slow against Google and Microsoft which are slaves to the evil empire.

Why do we resist being disarmed and let our government have a monopoly on munitions and armaments? Is there a better way to make us sitting ducks?

Put into context the AI being our source of official information in the future. Is it going to let us know things better? What will be its narrative on Ray Peat and the bioenergetic principles he espouses?

And how do you propose we stand up against it? Is it better to bury our heads under the sand and chant "No to AI in whatever form and shape. Pure pure evil."

Is there a context here that deserves to be considered or are we living in a vacuum of idealism?

[ThinPicking]

@Peatful said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

I miss him

🧡

[ThinPicking]

Excellent. To be continued!

(Others chime in.)

[sharko]

Friends, to all naysayers: you are going too far with your thinking.

I use AI to get answers faster and what I do today to help myself and others with health, I could do maybe in 5 years of intensive learning without AI.

I do not refer to all the beautiful theories because they are not relevant to my reality. In reality, to get real benefit from AI, you have to learn to work with it correctly. Just like the difference between one person and another person searching for answers on Google.

For example, using AI to get an answer to a question like "How to treat anxiety?" And applying his answers is as bad as searching on Google and acting on the first result. It uses me as a much more effective and faster research tool than Google.

And all the naysayers with the beautiful theories: I understand that you don't use search technology to learn. Do you only use old books that gather dust?

[Kvirion]

@yerrag said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

I have for now only my own thought processes to guide me in my own expectations from AI. I can only think of it as extremely useful as uncovering concealed dots which I have overlooked, leaving me to spend more time connecting the dots in search of solutions where cause and effect is brought back from exile in the current orthodoxy of academia and science.

I'm with you. I often use peatbot to find extra info. However, it is worth emphasizing that LLM can also serve us faulty dots...

[yerrag]

@Kvirion said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

@yerrag said in Bioenergetic AI Labs: The future of Ray Peat's legacy is already set:

I have for now only my own thought processes to guide me in my own expectations from AI. I can only think of it as extremely useful as uncovering concealed dots which I have overlooked, leaving me to spend more time connecting the dots in search of solutions where cause and effect is brought back from exile in the current orthodoxy of academia and science.

I'm with you. I often use peatbot to find extra info. However, it is worth emphasizing that LLM can also serve us faulty dots...

Good to be reminded of it as always a work in progress.