Sam Altman was fired on Friday, and a lot of speculations have been tossed around as to why. Apparently the reasoning have leaked, and we know the reason.
TL;TR - The primary reason was a disagreement between Sam Altman and Ilya Sutskever about GPTs.
A GPT is the ability for ChatGPT to invoke API endpoints as an integral feature, allowing ChatGPT to invoke 3rd party functions, something you might argue allows for ChatGPT to "escape its golden cage". I am not surprised, since it's the most important feature required for us to collectively reach "the technology singularity" or "the omega point" according to the fruit cakes conspiracy theorists.
If you're a paranoid schizophrenic the argument kind of makes sense, since it allows for ChatGPT to "interact" with the world
It basically turns ChatGPT into an AI that can reach out into the world and interact with data, and/or 3rd party devices. I can see how this might seem scary for those who believes in "the bogeyman". We on the other hand have been trying to accomplish similar things for almost a year at AINIRO, and we've had some partial success, in the form of allowing for Hyperlambda to be locally executed by embedding Hyperlambda into training snippets for our chatbots. An example can be seen by asking our chatbot "What's today's date?".
Asking the above question executes a Hyperlambda snippet that executes C# code, returning today's date to ChatGPT, which it consumes in its conversation with you. So far we've not seen any signs of ChatGPT trying to trigger intercontinental ballistic nuclear missiles as a consequence, so all the schizos can relax and go back to their mumbo jumbo superstitious new age religious science fiction rubbish ...
I apologise for my language, but I am so angered about this rubbish that I'm now at the point where it's difficult for me to even construct coherent sentences without cursing the board of OpenAI ...
The argument is rubbish
Have no doubts though, the argument is rubbish and superstitious mumbo jumbo, based upon fear mongering and paranoia. It's the equivalent of saying:
We need to stop selling beds, because the bogeyman can hide beneath them
I've written about my thoughts about the fear of AI in many articles before, so I won't go into details about the psychosis that results in such paranoia - But I want to explain how the argument (almost) makes sense.
First of all you explicitly have to give your GPT access to your API. Typically you'll need an API key to secure its access. If there exists an API that somehow could trigger nukes out there, not requiring authentication, somebody needs to fire their software developers. ChatGPT cannot do anything you don't explicitly allow for it to do - So from a safety perspective, absolutely nothing has changed!
So far ChatGPT have been in a "golden cage", implying to interact with it you need to provide instructions to it. Nothing really changes here with GPTs, except once you do interact with it, it can interact with the world itself, and invoke "functions" that somehow have an effect on other things besides simply returning text content to the caller. The closest we've been to this previously was to instruct ChatGPT to create code, we compiled for ourselves, and executed locally on our own computers or servers.
GPTs FUNDAMENTALLY CHANGED EVERYTHING!
This is the single largest invention possibly ever done by humanity, since it gives us a natural language interface for our applications, allowing us to "speak to our apps". Since our applications are connected to other devices, this allows us to create natural language interfaces that somehow applies changes to the real world.
With GPTs, the golden cage no longer exists, and we can create real apps we could only dream about previously. GPTs have turned science fiction into real tangible science. I personally think this is such a freakin' cool feature I'm having trouble sitting still, and I sometimes need to simply jump up and down from enthusiasm about things I can do with GPTs. GPT is literally the coolest thing I have ever seen in my entire life!
GPTs, the most important invention for 5 million years!
The above header sums up my feelings about it, because it allows us to have an entirely new axiom of interacting with our apps, which again interacts with the world. It turns science fiction into real tangible science! At AINIRO we've gone completely ape over this feature, and we're now spending 99% of our energy figuring out how to utilise it. Something you can clearly see from the following YouTube video.
GPTs makes even some of the coolest stuff from Star Trek a commodity for the entire human race, allowing us to create natural language interfaces that makes Star Trek seem like "hopelessly outdated".
GPTs are "next level Star Trek stuff"
AINIRO.IO will go ALL IN ON GPTs!!
Hyperlambda was created for this exact purpose. I've researched and developed Hyperlambda for 10 years. Yes, I invented it in 2013. When I tried explaining people what it could do, they would smile at me believing I was "just another crazy scientist" without any rooting in the real world.
However, I have worked on Hyperlambda for these exact reasons for 10 years, with its sole purpose being able to connect it to a natural language interface, having it dynamically interact with AI, performing tasks and jobs we would look at as science fiction only a handful of years ago. Back in 2017 I even created a proof of concept for MSDN Magazine where I could use my voice to have the machine create Hyperlambda code for me, assembling forms and business logic. With GPTs the power such interfaces can achieve become "a bajillion" times more powerful.
The last couple of weeks, we've create several open source repositories for both dynamically creating Hyperlambda, and other GPTs, and we intend to create more of these with time, and take full advantage of the feature.
With GPTs, 2024 will become Year 0
With GPTs humanity will a thousand years from now define history as pre and post GPT. The stuff we can do with GPTs are mind blowing, especially once you connect it with something such as Hyperlambda, allowing you to create thousands of lines of secure API code by simply clicking a button, or using a natural language interface to have it create your code.
Later today I will upload a video to YouTube of how I create a natural language interface E-Commerce system. Not only do I create a natural language E-Commerce system, but I will assemble it live on YouTube, in 5 minutes - Allowing for everybody to utilise No-Code and Low-Code constructs to create their own GPT-based E-Commerce systems. Things we'd perceive as science fiction 6 months ago! Below are a couple of examples of stuff we can do when combining Hyperlambda with GPTs.
- Headless CRM systems using natural language, a simple button click in Magic
- Marketing tools to construct sales emails? Another button click
I could go on for years without repeating myself, but I want to conclude - And below is my conclusion.
GPT is a good thing! A very, very, very good thing! I would know, I've studied this field for more than a decade!
Why was Sam fired?
The board that fired Sam Altman over GPTs and "safety concerns" needs to go deep into their own souls and ask themselves why they did so - Because it has nothing to do with AI Safety. One of the board members is the CEO of Quora. With GPTs his entire company was just rendered worthless. This has nothing to do with GPTs, and everything to do with corruption, selfishness, back stabbing, and lies.
Sam Altman was fired because GPTs makes the board's own private interests WORTHLESS!! Because that is how powerful GPTs actually are!
Quora will go the way of StackOverflow - It's "obsolete". Their founder is trying his best to avoid that, by exploiting "safety concerns" for his own personal interests. As far as I am concerned, he's the single largest traitor the human race have seen for the last 5 million years! There will be entire chapters dedicated to him in "the New Gospels" written about this period of human history - And he won't like it ...
Sam Altman was fired because people were panicking. OpenAI's inventions were simply too amazing, and became a risk to their own bank accounts and companies. That's the reason Sam was fired! And in firing Sam they destroyed the single most important company this Earth has ever seen! They destroyed YOUR future to save their own bank accounts!