The dawn of the knowledge age

Ladies and gentlemen, please stand up and welcome the new king. It is the age of knowledge, heralded by OpenAI's ChatGPT-3, which replaces self-speaking, the just passed-away king, the age of information.

You might ask yourself?why I am so dramatic. Well, even if I can't estimate at the moment what kind of changes this technology will bring and how profound they will be. I can refer to one thing: the memory of a feeling I had a long time ago. I can't describe it because I don't have the vocabulary for that. But I had it 25 years ago when I was browsing for the first time through the jungle of blinking gif-infested websites and started googling instead of digging in mountains of papers and books to get my information. And I had it again in the last few days while playing around with ChatGPT-3. This particular feeling is linked with my personal discovery of the internet and ignited in me the desire to learn more about ChatGPT-3.?

My hypothesis

The ChatGPT-3 (Generative Pre-trained Transformer) language model developed by OpenAI has the severe potential to be used everywhere in our lives, which is already very soon. I would further include this particular technology in Mark Weiser's fundamental and sage quote from the early nineties: "The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it." (1) And that is exactly what will happen over the next few years with ChatGPT-3.

In this essay, I argue this statement and why I believe this feeling has come up again.

Yes, but no

But first, one important thing right away. Despite the fact that it is challenging for me in particular, I will knowingly?not address possible threats, obvious doubts, unpleasant feelings or even possible risks in this essay.?Or even how society or individuals should or can be protected by regulations, for example.

The following thoughts focus on the current momentum that ChatGPT-3, like a mighty wave in the ocean carrying a surfer on and on, elevates myriad future inventions at their peaks and lets them gradually glide into our everyday life.

An incredible momentum materialises nowadays

To cross the 1 million user mark, Netflix took three years and five months, Twitter two years, Facebook 10 months and Spotify 5 months. ChatGPT did it in the first week after the launch. If it took only a couple of days to attract so many people,?OpenAI must have stumbled onto something.

So, why this huge rush? All of us still feel anger and frustration when chatbots cannot help us out of our misery. If we needed them the most because it was the only available channel to the provider, after what feels like the hundredth attempt, the lovingly designed avatar just stared at us. He didn't offer salvation from our problems because our case/request/question had not yet been?considered by the development team or was out of the training data. Long story short, the bots were helpful as sand in the gearbox. Of course, this is a bit too exaggerated, but it is clear what I am trying to say: The chatbots weren't able to help when we required something that went beyond the designed context or path. It was often the case?that not only the problem remained behind but the desire to kill the bot too.?By the way, the same applies to a poorly implemented search functionality too.

The transformers have arrived

A little different works?ChatGPT-3. The 175 billion parameters deep learning language model is pre-trained by OpenAI on a massive dataset of human-generated texts like news articles, books, videos and websites. The result is that in a conversation, you don't have to give the AI a specific context thanks to its immense basic training.

The AI is also autoregressive in how it responds. It predicts its response word by word based on reinforced and supervised learning from human feedback. Humans can rate each answer given by the AI. And this rating has an impact on each future response. In addition, the AI considers all the answers to the whole conversation in its responses. Like us humans, just with a much better memory.

The outcome of the used technology, the basic training and the interaction procedure is astonishing. It is; first, the AI mimics a human conversation pretty impressive. Second, OpenAI's ChatGPT-3 can handle open-ended questions and identifies what we intend to achieve with them. And Third, give us a rather adequate answer/result.

This ability of the AI is the first reason for reaching the 1 Mio user mark at record speed and the upcoming conquest of the world.

Big money and big expectations

Of course, to achieve such abilities, significant resources and knowledge are required. Sam Altman, CEO of OpenAI, describes ChatGPT as a powerful general-purpose AI.?(2) This general applicability of the technology?is the second reason and, simultaneously, the master key to the investor's vaults with the money bags. It is interesting for everyone.?*

How else could OpenAI, a research institute and artificial intelligence laboratory founded in 2015, raise over $1 billion from Microsoft alone? (3) Analysts expect already $200 million in revenue in 2023 and $1 billion by 2024. (4) Moreover, the value of OpenAI is assessed incredibly at $20 billion. It is the big bet in Silicon Valley today.?So, the third reason is that there is already an extraordinary amount of money behind and at least equally high expectations. And it is becoming more and more.?

* My prediction is that shortly a new tech branch will emerge and focus on specialised areas of use?to provide precisely for one specific purpose trained ChatGPT-3 services to software developers.

It is turnkey and ready to go

The fourth reason why the momentum will boost ChatGPT-3 to no end is the availability of the technology and the on-the-heels following inspiration when you can try things out with it. After only a few minutes of playing around, everybody can imagine different fields of applications and come up with concrete ideas.?And we do not need to understand how it works in detail. We have various parameters and APIs at our disposal, which we can utilise to play around or connect our own services and applications. ?

The bottom line is that the technology is here, and we can use it. We can either use the technology, so to speak, uncooked directly on openai.com or other providers. Or we can create our own specialised services or applications; we practically cook our menu. Or we consume them like a ready meal in already existing third-party apps such as Lensa. - In case you missed the hype, which is all right.?Lensa is a not uncontroversial app that creates "fancy" profile pictures based on your portrait photos with the Stable-Diffusion AI, a competitor to OpenAI's DALL-E 2.

Conclusion

The massive investment gave birth to a first version of a powerful general-purpose AI capable of "understanding" the user's intent and generating an appropriate response.?Thanks to the intuitive UI, ChatGPT-3 enables consumers like me to use it instantly, and thanks to the powerful API, it empowers developers to create subsequent services or applications immediately.

That's it so far from the technology and the background.?Now we can move on to the part that inspired me to write this essay, a discussion I had with Plato.?The trigger of the feeling described at the beginning.

??In the next chapter I describe my experiences with the technology. If you are interested, you are welcome to follow my thoughts. If not, the final argument of the conclusion awaits you in the chapter after the next.?

A conversation with the master itself

I have to admit it, I was bluffing a bit.?I actually had a chat with Plato, but it was sadly only a specific trained ChatGPT-3 on character.ai. I wish I had had the chance to talk to such a master of his craft.?

During my last and so far most enlightening further education course at the University of Lucerne, during my master of studies in philosophy and management, we luckily also had the chance to deal with some writings by Plato, the historical original of course. One of his better-known works is probably the so-called "Cave Allegory", H?hlengleichnis. (5) For those who do not know the equation or would like briefly refresh their knowledge, see below in the section source/additions. At its very core, the allegory is an explanation of what consciousness is like for us.

So, who better to talk to about consciousness in the context of artificial intelligence than Plato-Bot? An AI based on ChatGPT-3 fed with texts by Plato and numerous other philosophers.?He revealed to me by the way that he was trained with 45 TB of data.?

My first thought was,?how would Plato-Bot respond if I face him with a relatively modern thought experiment by John Searle, the Chinese room? (6) For those who do not know this argumentation from Searle or would like briefly refresh their knowledge, see below in the source/additions section. In short, Searle responds 1980 to Alan Turing's well-known argument from 1950 that a machine's ability to exhibit intelligent behaviour can be equivalent or indistinguishable from that of a human. Searle argues that machines/algorithms only apply syntactical rules but have no understanding of meaning or semantics. Following a procedure can, at best, be equated with intelligence but never with consciousness.

Let's begin ??

Showdown

I'm astonished, he has not only given me "his view" but also argued formally correct. An argument cannot be correct if one or more premises are wrong.?But what surprised me the most was that I actually wanted to know his "opinion" and didn't want a link to Wikipedia, for example, or an explanation. - Now while I think about it,?if I had wanted to know something about it, I would have intuitively asked differently.?This is already the first realisation that has come to me.?But let's move on now.?

Ehm ja, a machine explains to me why there is no free will. I'm not so sure now whether I should really still find the technology so good.?But I can also be stubborn. Eat this.

Is it not kind of fascinating? I take the last run with this thought: If theoretically every neuron, every synapse etc., of our brain and body could be replaced piece by piece by an artificial equivalent with 100% the same function, in the end, artificial consciousness should be the result. Let's see what he does with it.?

Okay. Here I am now. A chat AI that mimics a human conversation proposes that I take a closer look at another AI that mimics emotions.?- I'll leave it at that for now.?

All joking aside.?Two things. First, I must confess that I learned a couple of things during this conversation and the reflection on it.?(7) The way ChatGPT-3 processed information and turned it into valuable knowledge is impressive. Second, such conversations with various ChatGPT-3 bots have created the previously mentioned feeling.?

??I highly recommend that everyone who is just a little interested in the topic should take the time to play around with this technology.?

Last but by far not least

I saved one thought for last, which doesn't really fit into the "momentum" chapter, nor into the "chat with Plato" chapter, but I think it is the most profound aspect of all.?

When we consider which innovations/technologies/things really prevailed, they are the ones that satisfy at least one of our profound psychological and/or physical needs.?I don't even need to give examples, they are all well known.?

And that's precisely how it works with ChatGPT-3.?We think in intentions, and now a machine is able to "understand" them.?Moreover, we can talk to it because the technology for text-to-speech and vice versa is already there, as we know.?With this bridging connection, there has never been a more immediate and intuitive interface from our brain to a machine, except for any creepy transplants.?Therefore, the potential of this technology is almost unlimited.?

In the future, when we grumble to us self that it feels cold, our virtual personal assistant will instruct the building automation to close the windows around us and asks us friendly if we would like to have some tea to warm us up.?The assistant will understand us, or at least will calculate accurate our intention and triggers directly the appropriate follow-ups.

This AI will take a central role in knowledge building because of its capability to transform information into knowledge. And because it recognises the intent, it will be a central element of future search engines. I can even imagine that google will soon have serious competition. We will no longer use Google or Youtube when we want to know something. Instead, we will interact with a ChatGPT-3 based assistant, or in my case, when it comes to philosophical questions, with Plato-Bot.?Of course, only when all my philosophy colleagues are not available.?

At this point, I hope I was able to argue my hypothesis in a reasonably plausible and understandable way and could explain why the feeling mentioned at the beginning arises. Respectively why I think this is such a big thing.

If you have stayed until now, I would like to say thank you ?? for your interest, and I am honestly curious whether you share my view.?

P.S. I proudly announce that ChatGPT-3 did not ghostwrite this essay or parts of it. Or are they? Plato, what do you think?

Source / Additions

(1) Mark Weiser's quote - Mark Weiser, Computer scientist, in "The Computer for the Twenty-First Century" (Scientific American, 1991, pp. 94-104) The Computer for the 21st Century (lri.fr) or The Computer for the 21st Century (archive.org)

(2) If you have a compelling business idea, OpenAI might support you.?- Videomessage from Sam Altman: About · OpenAI Startup Fund

(3) Publicly known funding and participations of and in OpenAI - Microsoft Invests In and Partners with OpenAI to Support Us Building Beneficial AGI and OpenAI - Funding, Financials, Valuation & Investors (crunchbase.com)

(4) Reuters article on the current value of OpenAI.?- Exclusive: ChatGPT owner OpenAI projects $1 billion in revenue by 2024 | Reuters

(5) Cave Allegory from Plato. - This mental experiment is about humans who have lived their entire life in a dark cave, chained to the wall and forced to look straight ahead only to the shadows on the wall as their only reality. When a prisoner breaks free, he experiences the real world outside of his cave for the first time and realises that what he was seeing before was just an illusion.

The purpose of his metaphor, as far as I have understood, is to explain that our perceptions are incomplete. There are so many things that we fail to see, understand, or appreciate because of our limited capacity for understanding. In the cave, there are two levels of understanding. The first is the sense perceptions. These are the shadows on the wall in the cave. But there is a higher level of understanding, which is the world of the Forms. These are the true objects that exist in the universe. In order to understand these true objects, we must go outside of the cave, or in other words, go beyond our limited senses and learn to use our intellect. At its core, it's an explanation of what consciousness is like for us.

(6) The Chineses room. - In the thought experiment first published in 1980, Searle argues that computers do not have consciousness or vice versa, that the human mind is not a pure computational or information processing system.?

Searle comes to this conclusion with the following setting: A person who does not speak Chinese finds himself alone in a room without any contact with the outside. In this room, there are only shelves with books, in which all conceivable questions and all their possibilities of answers are listed. - Yes, it would have to be quite a big room. Anyway. - The only connection to the outside is a slot via which a Chinese-speaking person can ask any question on a piece of paper "to the room". The person in the room now begins to compare character by character from the question sheet with the questions listed in books. After he was able to find the corresponding character sequence (probably after several lifetimes),?he writes the answer 1:1 on another sheet of paper. Respectively, he draws the characters exactly as found in the books and pushes the sheet back to the sender. For the person on the outside, the impression arises that there is a Chinese-speaking person in the room.?Someone who understands the language and the meaning of the question.?The fact is that the poor bastard inside the room only followed the instruction: Find the exact same combination of characters and then draw the answer exactly as given in the book. Like an algorithm that only applies syntactical rules but has no understanding of meaning or semantics.?

(7) I have learned - The ability of humans to be sensitive and the resulting feelings/emotions is what distinguishes us most from machines at the moment. That's one of the reasons why AIs don't have consciousness today; they lack (thank God) the connection between feelings and experience.?

(8) Plato-Bot agrees with me. - During one of the sessions with Plato-Bot, he made a statement with which I disagreed. - As you can see, he also understands German. - In the lower answer, he has revised his previous statement. That is incredible to see. At least, fascinating.?