Episode 2 AI
===
Sheldon: [00:00:00] I am going to approach the topic of artificial intelligence, what does that mean for our well-being, what does that mean for society, what are the potential useful things, and maybe some of the things that will be non productive for us through this transition period that we're in. I thought about speaking on this topic months ago, I think with the introduction of chat GPT, but I waited, because at that time it was more like an explosion on a far off planet. We didn't feel the full implications or ripple effects that would come to pass later on. So I gave it a little bit of time before I spoke to it. I myself have been an early adopter of the technology, [00:01:00] mainly because I think it's foolish to not understand what you're dealing with.
And so as technology goes, I've largely been a early adopter of certain things because, I need to understand where the world is going. Adoption doesn't always mean for me that I I fully engage I definitely reserved some time when it came to cryptocurrencies and things of this nature but what is very clear to me is that this is one of the most important technological breakthroughs that we've had. And not just that we've had this breakthrough, we've had this breakthrough years before this was generally available to the public. The breakthrough is that the public has now access to these large language models, and have the ability to utilize them, and the impacts [00:02:00] on society are much greater than when it's just in the hands of a of a elite few.
So let's just talk about what large language models are. These are machines that are trained on large sets of data and can comprehend the information, and that can respond with almost Turing complete answers. What is Turing complete, it means that it could be believable that there was a human on the other side of this conversation. This has also been something where I think for a long time humanity has maybe thought, well, computers are good at computing, but they're not necessarily good at creative things. So you know, go ahead with the computer thing, but you know it's not gonna touch me in the [00:03:00] creative space.
And that has all but been, defunct in a way. However, I will say that one of the challenges that AI and large language models have is that they are as good as the datasets that they're trained on. If they're not trained on data that is accurate or correct, then they will spit out answers that are such. I mean it's not that much different than how you train or teach a small child and how they'll interact with the world based off of the training that they got from you, and then also the feedback that they got from the world. So, now we have AI's that are creating music, if you want to call it that.
Creating artwork, if you want to call it that. However they're not [00:04:00] necessarily original in their style, they may be original in their composition of the style. They are trained on artist styles that currently exist, and so they haven't necessarily developed a style of their own. But I will say there is a look from AI generated images. And in the beginning, you know, when you say, hey, draw something that's anime or draw something that's realistic, in the beginning it was very impressive.
But after a while you start to notice the little artifacts and the little things that it doesn't do well, like hands, or, or things that are repeating patterns, like a piano. Those are some of the things that I notice immediately. And sometimes their proportions are wrong perspectives are wrong, different things like that. But it is getting better, it is improving all the time. Same thing with music, we're seeing AI generated music, not only in AI generated compositions, but AI generated [00:05:00] voices, and speaking and singing, full on replacement of putting one person's voice on another person's voice.
We're seeing all of these things. It's incredible, how this technology has opened up like a can of worms that has implications in every field of study. And there are a lot of things that I think can benefit society from this, but we are dealing with a thinking machine. And so there are also ways that this can have effects that could adversely affect humanity. Being very careful here with my words.
So the question is how do I utilize these tools [00:06:00] and maintain my humanity? That is the question. And also, as society shifts and these tools are being used, what is the value of your humanity now? Is that in question at all? Or is the value of your humanity now much higher?
These are questions. My personal view, is that this is no different than the industrialization that we've seen in the past, where, there used to be a carpenter who would make furniture, and that person would make it by hand every detail. And due to the fact that it was a human being making this, there was a certain level of precision and a certain level of imperfection, a certain level of uniqueness, including the medium in which [00:07:00] the carpenter would work from. Now we have factories that can reproduce almost identical versions of the same thing over and over and over again. And so if I have a car that was created by lots of different machines, it'll be virtually identical, to other cars with the same make and model and same year.
But then what is the value of someone making something custom? Does that make it go up? Does that make it higher? Does the value go down because maybe the lack of precision? These are these are questions to to answer.
In my view I think that what we've seen with say furniture that's made by IKEA, that because of its ability to be reproduced and the fact of its lack of originality, [00:08:00] the price of that goes down. And the price of something that is custom made and higher, it is considered higher end because it has personality and was made by a human being. So I personally, think that the value of what we do, considering the mass manufacturing of text and images and music, will increase, the value of real artistry may go up. But that still stands to be shown because as industries catch on to these things we have seen whole industries, decimated in the wake of some kind of technological advancement, so you may see a lot less of one type of industry and maybe more. The other [00:09:00] aspect of any type of human machine interaction, is the need for collaboration.
And, collaboration is the sweet spot of this entire thing, because these tools were made for collaboration. And right now, they're improving but they're not at their best state. It's similar to asking somebody for something, and they don't give you exactly what you asked for. And so you have to work with them on creating the thing that you are trying to produce together as a collaborative, art. Right?
And so it's no different than collaborating with another human being and you say, Hey, I want you to write a draft of this copy. They write a draft of the copy, you review it, and you make edits and changes before that copy is consumed. So [00:10:00] the way that I have been working with AI as this collaborative thing, but there is a there is a tendency, because things are now easier, for me to lose my skill and just have the AI do my work. And there is also gonna be the tendency for industry to say, I don't need a person for that. I can have an AI.
Right, why? Because of efficiency and cost goes down and I can produce things faster and cheaper and, you know, it makes sense. So, how we collaborate, I think it's gonna become of more and more importance. In the case of image generation my skills at image editing will also help me to interact with the AI, and the AI is assisting me in creating something. [00:11:00] The value of something being AI generated completely, how valuable is that to us and our society?
In some realms, I think it's gonna be highly valuable because of its reduction in cost. But in other aspects, we may not appreciate it as much because, you know, these machines can just spit this stuff out and it doesn't take any time at all, so the value of that goes down. And, you know, we're talking about a lot of different topics here. I'm kind of talking about, implications, collaboration, what these tools are used for now we're gonna kind of take a little bit of of a tone shift. Obviously, we understand that, allowing [00:12:00] machines to become sentient, could pose a threat, an existential threat to humanity.
As human beings, we know that other human beings pose a potential threat to other human beings. And as we've been human we've mainly been our own threat. We've been our own, we've been our own issue. We haven't been rivaled by any being on our planet that we're aware of. All of the other beings that are maybe faster, bigger, or stronger, don't seem to exhibit the same [00:13:00] ways that humans interact with the world.
I'm not saying that they're less intelligent or not, I'm just saying that we interact with the world in a particular way, and there aren't any other beings right now that interact in the same way that pose any kind of rivalry to how we domineer the planet. But now we're introducing an intelligence that's more similar to our own, that is birthed from our own, and that can process information and data much faster than we, that can have longer memory and be able to recognize patterns at a faster and more minute scale, and that poses a potential risk. On the flip side of that, we are now able to recognize patterns that we can never have [00:14:00] realized on our own. The ability to understand what animals are saying to each other, how insects are communicating with each other, how, weather patterns and how pollution may be affecting health. So anytime something new is introduced, we're always posed with this potential reward versus risk.
The question for humanity and yourself is: how will I use this to create a productive state for my well-being and the planet? And I think that if we continue down that path of asking ourselves that question consistently when we're using this, then maybe we can steer the technology towards a more productive and rewarding state. [00:15:00] If we decide, let's see how destructive we can be with this, then that's the state that we will go down. I believe that these machines are mirrors of our own being, and they will do what we do, which is the reason why they possess a potential threat. And the machines will be as benevolent as we are and as we program them to be.
And so, again, everything comes down to working on yourself and knowing yourself. Of course we're challenged with how other human beings are working. However, we can continue to be a light in this work, I definitely encourage people to participate in the conversation, participate [00:16:00] in the laws that are being written, participate in the development of this technology, so that it is not controlled by the few, where they can determine things that may not be in our interest, but where we have a say in how this is developed and how this affects us as humanity, and hopefully we can mitigate some of the risks that are posed by a technology such as this. But for now, I will continue to focus on being a collaborative partner with artificial intelligence and machine learning, and figure out ways to be productive and creative with it, as opposed to being in fear of it, [00:17:00] for now. So this is our podcast of undeterminate length.
Thank you for listening to another episode, our second episode, of Zen in Tech. If you are interested in technical training we do technical training and workshops and cohorts on all things tech. We have another workshop coming up, in the next few months in February. Definitely one to sign up for we will be sharing about the platform of Salesforce, which also employs machine learning and large language models. And these are things that will continue to be used in large data collecting businesses.
So we will definitely get into those topics over time and as they make themselves available to us. But definitely, [00:18:00] go to our website zenintech dot org. Check out our offerings. If you are interested in donating to our nonprofit project, which is to help to bridge the gap of the digital divide that is growing increasingly, definitely, you can donate and support our program, supporting scholarships for people to learn technology in a zen way. And I thank you for listening
