Leo Does Political Philosophy With Advanced AI

[AstralProjection]

@Leo Gura You should make your own GPT and release it to the public so that people can ask you infinite questions. LOL

[Bobby_2021]

4 hours ago, Novac08 said:

@Bobby_2021 That's a beautiful song

Check the comments 💔

[trenton]

I have been using chat gpt 3.5. If there is a huge difference between this and gpt 4, then I need to check it out. GPt 3.5 is terrible at chess and I don't find it helpful at all.

[Bobby_2021]

1 hour ago, trenton said:

I have been using chat gpt 3.5. If there is a huge difference between this and gpt 4, then I need to check it out. GPt 3.5 is terrible at chess and I don't find it helpful at all.

Making these models play chess is a good way to see how good they are in solving a general task. Which they fail miserably at. Simply because there are too many variables to consider to make a decision.

Models specifically made to play chess can do that, but they can't be good at other tasks.

As far as the models are concerned claude 3 opus is by far the best. Beats gpt4 by a long shot. Gpt4 is a sad case of the corporate bending down to the whims of the normies. 

You can use perplexity AI for 20/mo and use gpt4 and claude 3 and some other models for the same fee. 

perplexity.ai

[martins name]

I was struck by how wise it was. It was like it had been told to study Ken Wilber extra hard. Then it struck me that maybe Leo prompted it to be wise by asking wise questions. Bringing up spiral dynamics for example will make the AI think of Ken Wilber for sure. I wonder if it has the same level of nuance if you ask it dumb questions. I can't use it as it's not available in Sweden.

[Sincerity]

On 3/31/2024 at 10:00 PM, Leo Gura said:

Let me know what you think:

https://www.actualized.org/insights/leo-does-political-philosophy-with-ai

Where do you think you get more perspective/food for thought from? "Human" replies here or conversation with an AI? Is there use in even asking us?

I think @Exystem's commentary was great.

I read through the document and it's really impressive but reading an equivalent of "you make a great point" for the 100th time made my eyes roll lol. The potential for getting stuck in an echo chamber is big. Unless you deliberately ask it to provide counter points for you, but that's not what you did here.

Edited April 2 by Sincerity

[Leo Gura]

1 hour ago, Sincerity said:

I read through the document and it's really impressive but reading an equivalent of "you make a great point" for the 100th time made my eyes roll lol. The potential for getting stuck in an echo chamber is big. Unless you deliberately ask it to provide counter points for you, but that's not what you did here.

Well, this conversation was just me testing out what this AI is capable of.

I would certainly make refinements to my process in the future, telling the AI to be less agreeable.

Using AI well is a skill. Yes, it can turn into an echo-chamber if not used carefully.

It's important to explicitly ask for counter-examples. I did some of that, but I could have done more.

Edited April 2 by Leo Gura

[Sincerity]

7 minutes ago, Leo Gura said:

Well, this conversation was just me testing out what this AI is capable of.

I would certainly make refinements to my process in the future, telling the AI to be less agreeable.

Using AI well is a skill. Yes, it can turn into an echo-chamber if not used carefully.

It's important to explicitly ask for counter-examples. I did some of that, but I could have done more.

Sounds good 👍

[Yimpa]

26 minutes ago, Leo Gura said:

Using AI well is a skill. Yes, it can turn into an echo-chamber if not used carefully.

It's important to explicitly ask for counter-examples. I did some of that, but I could have done more.

That’s one of the major benefits of ChatGPT, the ability to set custom instructions

 https://openai.com/blog/custom-instructions-for-chatgpt

[Davino]

20 hours ago, Davino said:

What other topic are you interested to discuss with AI?

I would like to see some discussion about comparative Religion. What is Religion? Why does Religion exist? How religions are similar or different among them? What is the purpose of Religion? How does religion evolve with maturity and development? and so on

Edited April 2 by Davino

[OBEler]

Claude 3 has an IQ of 101. So a little more intelligent than most humans:

https://www.trendingtopics.eu/ai-ranking-according-to-iq-claude-3-beats-chatgpt-4-and-co-in-intelligence/

Absolutely crazy .

For me AgI is not far away 

Edited April 3 by OBEler

[Leo Gura]

9 minutes ago, OBEler said:

Claude 3 has an IQ of 101. So a little more intelligent than most humans:

https://www.trendingtopics.eu/ai-ranking-according-to-iq-claude-3-beats-chatgpt-4-and-co-in-intelligence/

It's way more intelligent than its IQ though, due to humans having so much bias.

Jordan Peterson probably has a pretty high IQ. But his bias cripples the intelligence of anything he says.

Chris Langan supposedly has one of the highest IQs in the world, yet his political takes are cringe.

The problem is that IQ tests don't measure bias nor cognitive development level. So you could have someone with 150 IQ who is a Nazi.

Edited April 3 by Leo Gura

[zazen]

IQ is having intellectual horsepower, wisdom is knowing which horse is worth riding on and in what race. Put another way; intellect answers questions, wisdom knows which questions are worth asking and getting answered. The power of AI belongs to those knowing how to use it - which is why Leo’s conversation was so insightful and fascinating.

From another thread:  Logic and intellect can bring smarts but not wisdom. To have a accurate map of reality and human nature requires humility and courage to face truths that may unsettle us. A mighty intellect can enlarge the ego which will then use the intellect to strengthen any delusions it may have and further rationalise to shield us from staring into reality.

Truth requires us to pursue perceived terrifying insights to their conclusions, that in the end may not be as terrifying if only we could get to them. Maybe this is where the warrior spirit of stage red needs to be integrated - the virtue of bravery and courage.

Without courage or humility, otherwise intellectual heavy weights are not strong enough to emotionally/spiritually handle their own intellects. They will instead build more elaborate systems of thought that are tightly self-consistent and logical in their own closed loop of delusion. Within their bubble of logic it all maps together cohesively, but none of it survives first contact with reality which it is divorced from.

While they exhibit logical thinking (a sign of intellect), the way they apply that logic is flawed and inaccurate, leading to foolish outcomes (signalling lack of wisdom). Cowardly self-delusion dressed up in fancy intellectual lingerie is still just cowardly self-delusion.

Edited April 3 by zazen

[OBEler]

22 minutes ago, Leo Gura said:

It's way more intelligent than its IQ though, due to humans having so much bias.

Jordan Peterson probably has a pretty high IQ. But his bias cripples the intelligence of anything he says.

Chris Langan supposedly has one of the highest IQs in the world, yet his political takes are cringe.

The problem is that IQ tests don't measure bias nor cognitive development level. So you could have someone with 150 IQ who is a Nazi.

Nazi leadership had all very high IQ, as was measured in Nürnberger process.

 

IQ tests results vary some articles claim ChatGPT 4 already has an IQ of over 140 in IQ test

[yetineti]

1 hour ago, Leo Gura said:

It's way more intelligent than its IQ though, due to humans having so much bias.

Jordan Peterson probably has a pretty high IQ. But his bias cripples the intelligence of anything he says.

Chris Langan supposedly has one of the highest IQs in the world, yet his political takes are cringe.

The problem is that IQ tests don't measure bias nor cognitive development level. So you could have someone with 150 IQ who is a Nazi.

@Leo Gura

I’m not trying to oversimplify, obviously bias, cognitive development, etc. exist.

However, bias, cognitive development, perceived intelligence, etc. are all tied together by knowledge, as well as intelligence.

AI’s seem smarter than us not just because they’re ‘unbiased’ - they have access to all of the information. 
 

Knowledge is not intelligence and vice versa.

Thats my ‘theory.’ Frankly I find their IQs to be around 100 still. I do not find these AIs smart. I find them knowledgeable.

Edited April 3 by yetineti

[yetineti]

Average IQ + all Knowledge = fantastical illusion these AIs are intelligent

[Bobby_2021]

8 minutes ago, yetineti said:

@Leo Gura. I do not find these AIs smart. I find them knowledgeable.

That's all there is to it.

Feeding these knowledge into a human mind and letting it make decisions would be really impressive. 

I am not impressed by AI, but it's sort of good. And AGI will not be there in the next 50 years. Don't fool yourself. 

[Sigrun]

Just finished reading it, what an amazing piece of technology.

It's a breath of fresh air to see it's not programmed to assume everyone is a moron who could interpret anything it says in dangerous and harmful ways.

What really impressed me though is how well it understood what you were saying and how sharp and relevant its responses were, as well as its eloquent use of vocabulary without sounding overly academic.

What shocked me was how it was able to understand social ques / humor better than most humans.

Besides that. the conversation itself was a pleasure to read even though I didn't get any big political / psychological insights from it since I my meta-perspective on what was discussed was almost all something I've already arrived to, but reading it discussed in such a sharp and eloquent way with basically no fluff or filler definitely helped solidify my meta-understanding.

Thank you for all your highly underrated contributions, Leo

[OBEler]

I think my recent Hitler thread started the initiation for Leo to use ai much deeper. Funny, after this discussion, I had the same idea like leo,  thinking about why not Leo uses ai more for self development and how much he would profit from it, helping him for his future work and bam...he then created an new blog post about exactly that.

 

[sholomar]

I love AI because it does such a good job at removing bias. It's no secret when I pop by here that I'm critical of much of the ideology, and it mostly comes down to a single paragraph of text in Leo's document: 

"The fundamental problem and damage with leftism is that it adopts utopian ideals without appreciating the stark and brutal survival conditions and the realities of human nature. When leftists underestimate the harsh realities of human nature, they come up with fantastical political ideas which are far too advanced for their time and lead to a clusterfuck if implemented."

My big critic of the left more than anything is their sense of moral superiority they tend to have about their attitude that almost reminds of me of a religion or cult, that vibe that comes off them... that those conservatives are so much beneath us, that we are superior...  it's a very off-putting vibe, from someone who tries to be totally without bias and doesn't identify as either "conservative" or "liberal" ... that said I'd acknowledge the conservatives would do the exact same thing if they had the power, and they have.  There is a definite bias here in the forum... yall just think you are right and they are wrong. Ultimately I would say it's more of a relative truth than an absolute one.

That said, I respect Leo, and all of you, and your position, and hope that humanity can some day reach your level of ideals, despite the evolutionary wiring of our DNA which resists it.  Personally I think genetic engineering is needed to tweak the animal instincts within us, particularly the fear response.... to selectively reprogram parts of our genetic software.

Edited April 3 by sholomar