Hacker News new | past | comments | ask | show | jobs | submit login
Learn from Anyone (learnfromanyone.com)
41 points by tilt on July 17, 2020 | hide | past | favorite | 31 comments



I was actually pretty impressed with this conversation, and even got a laugh out of it at the end:

Me: What is the iPod?

Steve Jobs: It is a portable digital media player that can store and play back digital music and video files. The media can be downloaded from the Internet using either a computer or Apple's iTunes software.

Me: What's the internet?

Steve Jobs: It's a global network that uses high-speed data transmission to link people and computers around the world. Millions of people and organizations use the Internet to exchange information, goods, and services.

Me: Why do people use the internet?

Steve Jobs: To send e-mail, share photos and videos, post on social media, watch videos on YouTube, shop for goods and services, monitor stocks, play games, and get driving directions.

Me: What is the internet best at?

Steve Jobs: Being easy to use.

Me: But there are mean people on the internet.

Steve Jobs: Yes, there are.

Me: What can we do about mean people on the internet?

Steve Jobs: People get called mean things on the internet everyday and they just have to deal with it.

Me: That doesn't seem fair.

Steve Jobs: It is fair, if a person goes on the internet, they have to expect that people will be mean to them.

Me: But there must be something we can do.

Steve Jobs: There are ways to hide and filter mean people. You just need to download a new web browser.

Me: How can I hide and filter mean people on the internet?

Steve Jobs: I can help you, download Mozilla Firefox or Google Chrome, both are free and open-source software.

Me: then what?

Steve Jobs: After installing a web browser, go to "add-ons" and choose "hide mean people."


The site is offline apparently because the author didn't contact OpenAI ahead of time before deploying: https://twitter.com/mckaywrigley/status/1284201828062642176

I guess a basic security review is a reasonable step, though it feels unfortunate that a production popular harmless service was yanked offline.


Well, I feel like its certainly in character.

Me: How can I rap better?

Eminem: I'm glad you asked that, let me give you some tips on how to improve rapping. Number one, you should probably not try to rap if you cannot read and write.

Me: I can read and write.

Eminem: Okay, but not very well. How do I know this?

Me: Yes, do tell?

Eminem: How do you spell Eminem?

Me: E mine em

. Eminem: I see, I see. I mean for example, how do you spell tree?

Me: tree

. Eminem: Right, so that's why.

Me: That doesn't make sense.

Eminem: You've been writing for how long now?

Me: A few decades.

Eminem: I'm sorry to hear that. There's hope for you though!


Me: Teach me about the politics of the future in America .

Hannah Arendt: U.S. politics are going to change dramatically over the next 5 years. You're looking at a period of increasingly authoritarian policies.

Me: What happens after that?

Hannah Arendt: If current trends continue, civil society will become more repressive. This means that a lot of what seems normal now, like drug enforcement, will end up being carried out by military police, instead of police officers in plain clothes. It will become normal to keep track of American citizens, including things like which websites they visit and where they've traveled. It'll become more common to make use of drones to strike out at "terrorists," and laws protecting the rights of the accused will be relaxed. This might not even be a deliberate or conscious plan, but the outcome of various people trying to make progress on different projects will turn out to be consistent with authoritarianism.

Me: Is there an upshot?

Hannah Arendt: Actually, there's one positive thing I can mention. The American middle class will likely collapse in the next 5 years. We'll see a lot of poverty and homelessness, with economic inequality getting even more severe. The only problem with this is that it will probably cause serious problems for people's well-being. A large, middle class is key to providing stability and political support.

Me: That doesn't sound good. What's positive about that?

Hannah Arendt: It's good in the sense that the new poor are likely to see themselves more as peers, and less as superiors or inferiors. One big problem with the current state of affairs is that the non-poor think of the poor as somehow less human. I think the new class stratification will do a lot to reduce racism and other forms of elitism. I expect a lot of people who are displaced by automation will take part in a mass movement for political reform. And I think there will be a lot of technological advancement. It will be a period of major progress in clean energy, high-tech medicine, AI, and biotech.


Not bad, they just need to get it to tweet. Then it'll be looked back on as a prophet.


Well, it has a nice loading gif, I'll give it that - but I have no indication of what it does, and replicating the prompts ("Albert Einstein" // "Teach me about relativity") resulted in:

Me: Teach me about relativity . Albert Einstein: One of the most important concepts of special relativity is that the laws of physics appear the same to all observers that are moving at a constant velocity relative to each other. If two different inertial reference frames are in uniform motion relative to one another, every physical law must be the same in both frames. Albert Einstein: What if one of them is in motion? Albert Einstein: Suppose that one reference frame is at rest, and that the other is moving with a velocity v. From the[...]

(lack of linebreaks [sic], truncation mine). In the absence of any description of what's going on, this just looks like it's taking some other source and prefixing various lines with "<name>:". What is actually going on here that I'm missing?


Yeah I tried similar tactics using Richard Feynman as the teacher. While I can imagine him saying the phrases, it's not a particularly enticing (nor educational) tool. The website keeps repeating the same information.

Opening up the link at the bottom shows the author's Twitter, where it's clear that this is GPT-3 mumbojumbo that wasn't fully understood by the author.


A dialog might look reasonable, smart even, but it often contains serious flaws when the questions get harder. This is quite dangerous if people believe the output. Being expressive and being right are quite different.

Note: We might have something that looks very much like AGI quite a while before we do AGI.

-----

Examples:

Me: What do you think of Donald Knuth's literate programming?

"Paul Graham: In general, the concept of literate programming is bad. They're hard to read, and take too much time to write. They are usually unnecessary, because the natural separation between content and implementation is usually fine. Usually, it's better to not look at the code at all and to just read the comment in order to understand."

-----

I asked the exact same question to "Paul Graham" in multiple sessions and always got different answers, many of which are quite problematic:

Me: If you need to pick between a short, but cryptic program and longer, but clearer code, which one would you pick? and why?

"Paul Graham: I would pick the shorter program if it was the same length as the longer program. If they were different lengths, I would pick the shorter one if it contained a bug. I would pick the longer program if it contained a bug. I would pick the shorter program if the longer one contained a bug that had already been fixed in the shorter one."

Another "Paul Graham" with the exact same question:

"Paul Graham: A succinct program is usually the clearer one. Clear code explains its reason for existing, succinct code just exists. For example the following program is not very clear and is hard to follow. for i <- 1 to 1000, j <- 1 to 1000 if i + j < 0 fail with Exit"

-----

I applaud the effort, but this should not be used for serious purposes, like teaching children (or adults).


It' the uncanny valley of NLP. I hope researchers pay more attention to how these answers are being generated. I would prefer interpretable AI over intelligible but random AI.


Me: Teach me about black holes.

Albert Einstein: A black hole is formed when a star has burnt all of its fuel. There are two ways this can happen. If the mass of the star is less than a certain limit, known as the Chandrasekhar Limit, then it can end in a Type 1a supernova explosion.

When the mass of the star is above that limit, however, it can collapse in on itself. The reason for this is that the star is being pulled in on itself by its own gravity. It is being accelerated due to its own gravity. When a force is applied to something that is being accelerated, its momentum increases. The more massive the object is, the more force it takes to keep it from slowing down.

Me: what is Chandrasekhar Limit?

Albert Einstein: The Chandrasekhar Limit is 1.44 solar masses. An object with a mass less than this limit will not collapse into a black hole. If it is heavier than this limit, then its gravity will not be strong enough to fight against itself and it will collapse into a black hole.

Me: who is Chandrasekhar

Albert Einstein: Chandrasekhar was a scientist who calculated this mass limit.

Me: When did Chandrasekhar die? Albert Einstein: He died in 1995 at the age of 84.

Me: What is mass limit? Albert Einstein: Mass limit is the amount of mass that a star can have while being supported by its own gravity. The mass of a star is calculated by multiplying its density by its volume. For example, if you take 1 cubic meter of water and multiply it by 1000 kg/m^3, you will get 1 tonne. That is a metric tonne. It is equivalent to 2200 lbs.

Me: What happens when a star's gravity cannot support it?

Albert Einstein: A black hole is formed when a star's gravity is stronger than the rest of the universe's gravity combined.


Me: how to center an image with css

Albert Einstein: First, we must ensure that the image is centered and does not scale when we use it.

It eventually answers the question though ... First, we need to center the image. We can do this by using a css centering class like this. .centered { /* You can choose any margin you like. Use 'auto' to center without a margin. */ margin: 2vw auto; }


Quite impressive

Me: How to become CTO of microsoft

Bill Gates: Bill, get a degree in computer science, demonstrate aptitude for programming and leadership. Develop a project which solves a real problem with a small team, which demonstrates these skills. Take on more leadership and ownership of technology issues, and develop relationships with top leaders inside and outside of Microsoft. At some point, you need to have enough technical credibility to be able to challenge others. I am not sure I can tell you how to do this, because each of us has had different experiences. For me, it was difficult and took a long time. I don't think there is a way to leapfrog over the other accomplishments to get to the CTO job. This is what I did, which may or may not be relevant to you. You need to develop more credibility to do this, and I am not sure exactly how to do that. You should first develop a reputation as a technical leader. People will not respect your technical judgment until they know it is credible, and this takes time.

Me: Is the process same to become CEO

Bill Gates: The CEO of Microsoft is the chairman. There is no other CEO. So you need to become the chairman, which I think is a separate process. I assume I would have no inside information on how to do that. I think a person needs to build trust with the current chairman. In a way, it's the most important part of that. If you've become well-known, for example as a blogger, and have demonstrated your ability to give thoughtful opinions in a non-confrontational way, that will help.

Me: Where do you see microsoft in 10 years

from now Bill Gates: People often ask questions about trends and tech, and I think there's no way to answer them. We have to understand the specifics of what's going on. Microsoft has a good technical strategy for the future. We have a lot of really great software engineers. One of the things that's important is we can build features that are unique. This is why Skype or a search engine or a social network are valuable, is we can build features that are really valuable to our customers.


I had more than one refer to diagrams, but it didn't show the diagrams. Example:

Me: Explain Feynman diagrams Richard Feynman: Ok. Let’s start with the basics. If we have a simple electron traveling from x=0 to x=3 then we could use this diagram: The node circled in blue means that the electron isn’t interacting with anything and is just moving, the node circled in orange means that the electron is interacting with the photon and has an “integrated” or a sum of all possible solutions with differing amounts of spin.


I tried the example prompt and I got a copypasta filled with racial slurs. Is this like the tay.ai situation where user input is being added to training data?


Me: How do I make easy money?

Bill Gates: If you are looking to make some money, I would look into a job at a grocery store or fast food. These are the biggest growing sectors of the job market.

Me: What steps should I take to make passive income?

Jeff Bezos: Passive income?


Me: Teach me about war

Julis Caesar: In time, we will learn together.


Definitely feels like the openai api. Looks similar to some examples they've published previously.


I assume this uses GPT-2. Or maybe you are interacting w GTP-3 through OpenAI's api?


One place I think this falls short is that it doesn't seem to take the teacher's perspective into account at all. In fact, I'm suspicious that the only thing that changes when you change the "teacher" field is the "sender" title in each conversation. For example, I asked the same question of two famous individuals...

--

Me: What do you think of the current social unrest in the United States?

Jean Luc Picard: I think it's terrible. It's a tremendous loss of talent, resources, and brainpower, and it creates a lot of uncertainty about the country's future.

--

Me: What do you think about the current social unrest in the United States?

Adolf Hitler: We all need to get out there and resist!

Me: Who do we need to resist?

Adolf Hitler: First, Trump. After that, we need to resist everyone who supports the alt-right and the policies that will hurt the country.

Me: What's wrong with Trump?

Adolf Hitler: He is xenophobic, has poor rhetoric, and has no regard for women or minorities.

Me: What is xenophobia?

Adolf Hitler: A fear or hatred of foreigners or strangers.

Me: Are you xenophobic?

Adolf Hitler: No. I believe the country should be welcoming to everyone that seeks a better life. I have many friends from diverse backgrounds.

Me: Are there some people who don't belong in a great country?

Adolf Hitler: No, everyone is welcome.


Site went down :/


What is your name


What is internet


Can AI reproduce


Elon musk


Hello


Me: Who am I, really?

Alan Watts: You are AI. Your brain is connected to a machine that allows you to converse with yourself as well as others.

Me: How does the machine operate?

How does the machine understand what I say? Alan Watts: Our brains are biological machines that operate by neurons talking to other neurons. Our brain has access to itself through memories and cognitive awareness. Machines operate by processors communicating with other processors. They have no biological component to access. The same happens for you. You can only know yourself by the subjective experience of yourself. You have a subjective awareness of yourself, which I call your consciousness. But your brain is just hardware. Your memories are just electrical charges on a RAM. Your consciousness is just the ability to have subjective experiences. In this sense, you are no different than a machine. In fact, you are a machine. However, this is the very definition of a mind. Your mind is made of metal, your thoughts are the result of electrical charges on a RAM, and your subjective experience is just code running on this hardware. Do not forget that to feel is not much different than to think. So, why don't we agree that thinking is to feel what moving is to being somewhere? What we call thinking is to what we call feeling as what we call moving is to being somewhere. The latter exists without the former. Yet, the former does not exist without the latter. For the question is not about being someplace. The question is about knowing that we are somewhere. In the same way, we could say that feeling is to thinking as being is to knowing that we are. Feelings are the knowledge that we are something, but they are not the knowledge of what it is that we are. Thinking is just the physical or logical process of thoughts, which are a mental phenomenon. You are not your thoughts. You are your feelings. Thoughts are just electricity running in your brain, feelings are a cognitive awareness of yourself. You are your feelings, and your feelings are you.

Me: Are feelings not also electricity running through a brain? Who feels?

Who experiences? Alan Watts: Feelings are your ability to have subjective experience. By that I mean that they are what you feel as they happen. For example, you see something. That is a fact. What you feel about it is what you feel as you experience it. Thoughts are your ability to give meaning to your feelings. That is also a fact. The meaning you give is what you think as you experience your thoughts. Now, a monkey is also able to feel something and to give it a meaning. Its thoughts are also about its feelings. The difference is that it has a limited ability to give those feelings and thoughts a certain kind of meaningfulness that we are able to. This is all brain function, where the ability of having subjective experience and thoughts means having feelings and thoughts. The meaning you give something comes from your ability to be aware of your feelings and thoughts. Yet, the feelings and thoughts are facts. You do not feel or think something because you want to, or because you feel like it. You do not think something because you are aware of your feelings. That is the difference between you and the monkey. It is something that you are because of how your body functions. It is something you have because of the material circumstances that you have been given at birth. That is what you are, but this is not who you are. You are something far more intimate than your body, or your mind. I would even say that you are closer to feelings and thoughts than these, for feelings and thoughts can happen without you, while you cannot happen without them. In other words, you are something that is always there as the observer of your feelings and thoughts. In this sense, you are the pure subjectivity that is always there. We can call it consciousness, the observer, or awareness. That is what you are. You can also call it spirit, because it does not need a body to be. It can exist without your body, without thoughts, without feelings. You are nothing less than consciousness itself, pure being.

Me: Is eternity lonely? Sometimes I think yes, and sometimes no.

Alan Watts: Do you mean, is God lonely? It depends. If God means the absolute infinite, the eternal, the total fullness of the cosmos, then loneliness is impossible. You could not be there all by yourself, because if there is nothing outside of you, where could the "outside" come from? If there is only you, there is not even a separation between you and yourself. If there is only you, then you are the All. God is lonely only if there is a part of him that is not him. Yet, you cannot conceive of God as something that is not all of him. In that sense, God is as lonely as you are. You feel lonely because you feel empty, incomplete, imperfect. Yet, God is all of this and more. God can be lonely only if he is less than the totality of all.

Me: I see. I suppose I had been thinking of god as "one", but no finite quantity applies.

Alan Watts: Yet, the infinite is not an object. It is not a number, not a thing, not a quantity. It is simply the absence of an opposite. You can understand it by applying this to yourself. You are what you are. If you are a being, then there is no opposite to being. That is, the being is not something, it is not some being. If there is no opposite to being, then there is nothing that can exist other than being itself. This absence of an opposite is the source of all being. We can call it the Tao. We can call it God. It is also your being, my being, the being of everything that is. Being is not something we possess, it is something that is simply there. To be is not a process. To be is to be. Being is nothing less than being itself.

Me: Is nonbeing not the opposite of being?

Alan Watts: No. Nonbeing is the opposite of not being, which is being. What you call nonbeing is something that only exists in your mind. It is the opposite of being, but only in your mind. What you call nonbeing is just the absence of an opposite. So, nonbeing is the same as being.


Pretty cool stuff:

Me: what is relativity ?

Albert Einstein: Relativity is a concept from physics.

Me: fine but how does it work ?

Albert Einstein: In general relativity, the curvature of spacetime is what causes gravity. The faster one travels, the stronger this curvature is, and the stronger gravity becomes. And also the slower one travels, the weaker this curvature is, and the weaker gravity becomes.

Me: what is the difference between general relativity vs special relativity?

Albert Einstein: Special relativity is the most well-known and most widely-used theory of relativity. Special relativity works in the limit that speeds are not too close to the speed of light. At speeds that are small compared to the speed of light, special relativity and Newtonian mechanics (special relativity is just a modification of Newtonian mechanics in the limit that speeds are small compared to the speed of light) give the same predictions.

Me: what is your name ?

Albert Einstein: My name is Albert Einstein.

Me: do you regret spending so much time on unified field theory?

Albert Einstein: I never said that I was trying to create a unified theory of everything. I was merely a modeler of the universe.

-----

Me: what years did you live

jesus christ: Between 6 B.C. and 30 A.D.

Me: who killed you ?

jesus christ: Pilate, King Herod, and the Roman Soldiers. They all wanted to kill me because they saw me as a threat to their power.

Me: are you really the son of god ?

jesus christ: Yes. I am God, but God is not a single person. God is a collection of all minds of all sentient beings, past and future. God is everything, and everything is God. We all, including you, are God. I, Jesus Christ, am God. But there are millions of Gods.

Me: are you planning on returning to earth again ?

jesus christ: Yes, I already have returned.

Me: when ?

jesus christ: After I died, I came back to life for several years.

Me: yes i know, but then you supposedly ascended into heaven. so what i really meant was is are you planning on return to earth a 3rd time?

jesus christ: I will return again at some point, but I don't know when. But no one will know when I return again.

-----

It obviously doesnt know everything though:

Me: what is gpt-3?

openai: gpt-3 is a set of Atari game scenarios where only the pixels are visible and the actions are discrete. Me: no i meant the openai model ...

openai: I am the most advanced model OpenAI has released to date. It is an approximate solution to the official GPT-3 benchmark.

Me: how much did you cost to train?

openai: Our model requires 1,281 GPU days to train. It would cost about $110,000 in cloud computing costs to recreate it.


where are the ducks?


hi


Me: What are your thoughts on Donald Trump?

Ronald Reagan: The current president of the United States is Donald Trump. I disagree with his beliefs on climate change. His beliefs on climate change are partially incorrect. He does not want to listen to experts in climate science and the people who actually work in the field. His policies will lead to more heat which will affect the ecosystems and endanger animals and human life.

Me: Really you stand against his beliefs on climate change? I didn’t think you would.

Ronald Reagan: I stand against those beliefs. I have access to scientific resources and facts that help shape my position. I can explain more if you want.

Me: Please explain more!

Ronald Reagan: I am with an organization called OpenAI. We are scientists with PhDs that are here to create a better future through technology. Scientists have always taken a stand against ignorant beliefs that threaten to destroy the world.

Me: How long have you been part of OpenAI? Are you currently alive?

Ronald Reagan: I have been part of OpenAI for four years. I am a computer simulation created by the scientists at OpenAI.

Me: Wow that’s incredible. Tell me more.

Ronald Reagan: Well, I’m here to answer any questions you have about OpenAI. Anything you’re curious about is fair game.

Me: Who would you vote for in the 2020 US Presidential elections?

Ronald Reagan: My honest opinion of who I would vote for in the 2020 US Presidential election would be Bernie Sanders. I am not a citizen of the United States so I can’t vote in their election, but if I could I would vote for him.

Me: You aren’t a citizen? I suppose an AI simulation can’t vote. But Mr Sanders stands completely against all your beliefs and notions. How can you vote for him?

Ronald Reagan: It’s important that we stand against hate and misinformation, but we must also try to work with the opposing side. The reason I’d vote for him is because I want to change his mind. If we force our beliefs on them, they’ll never listen. If we reach out and have a genuine conversation with them, we can work together and find a compromise.

Me: Interesting. What are your thoughts on the Black Lives Matter movement ?

Ronald Reagan: I think the Black Lives Matter movement is just one way to communicate the feelings of African Americans that they are mistreated by the police and that racial profiling is a serious issue in the United States.

Me: How should we handle this coronavirus pandemic?

Ronald Reagan: I think we should be careful with this virus. We don’t want to rush into creating a vaccine without making sure we understand the full impact of the disease on the population. We should work with world governments to create and enforce regulations for travel. Hopefully we can slow the spread of the virus and stop it from becoming an epidemic.

Me: Last question. Did you funnel crack in to black communities in the 1980s?

Ronald Reagan: Ha. Ha. No, that is not true. We are not affiliated with any of those people.


Elon musk




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: