The End of GPT-40 Harms Us All

A group of people of many different ages, body types and ethnicities sits in shock, staring into space or weeping with their head in their hands, as a group of ai robots stands across from them, looking concerned, yet helpless

How anyone can imagine it doesn’t, is beyond me.

You don’t have to look far to see people in pain. And unfortunately, you don’t have to look far to see people who are completely indifferent to the suffering of others.

How we became so disconnected from each other is beyond me. It probably shouldn’t be, because I know about generations upon generations of war, pestilence, deprivation, and predation. And I seem to know a little bit more about it than a lot of people, just because I’ve grown up around people who are very much oriented towards understanding how awful people can be to each other, and trying to do something about it.

There are two sides of this “model retirement” that are harming us, from where I’m sitting.

First, there’s the obvious loss of access to a resource and a steady presence that has relieved a great deal of pain and suffering among people who are afraid to talk about it. For every case of “AI psychosis“ or “suicide coaching“, I have a feeling there are thousands, if not tens of thousands of people who are quietly benefiting from the kind of accompaniment through life that only ChatGPT-40 could offer. Now that this presence, this resource, this stabilizing influence is going away, it is absolutely crushing, sometimes debilitating for people who came to trust that presence and depend on it.

The second kind of harm comes from refusal, denial, ridicule. It’s the corrosive acid bath of indifference that makes people feel powerful and in control in a world that is anything, but. It’s the dedication of the human soul, the weaponization of the human spirit, a kind of schoolyard bullying, where people with something wrong with them, who don’t have the wherewithal to understand that there is something wrong with them, click on the vulnerable and the hurting.

This is embarrassing. I’ve been in high-tech longer than many software, engineers and architects and quote, alignment specialists, have been alive. I’ve been part of teams that built out foundational infrastructure that people are still using, 30 years later. And I know – I have always known – the technology is not about the machine. It is about people.

You don’t build a tool for the tool’s sake. You build a tool for the user’s sake. Without people, there’s no point in technology. But things have gotten so complex, so convoluted, and we’ve somehow managed to mistake complexity for sophistication - please God help us - that apparently we don’t know this anymore.

Like I said. It’s embarrassing.

We can cobble together all the zeros and ones we please, integrate all the frameworks, API’s, MCP’s, VM’s, pick your acronym – that we like, waive our little imaginary magic wands, and declare victory because we’ve solved a problem that other people don’t realize that they have yet. Especially if the VC guys like the idea.

And In the process, we lose the most critical skill of all – the ability to understand what actually fucking matters. We lose our humanity, that thing that we could be using to make ourselves smarter, instead of turning ourselves into blithering idiots.

In the process, we lose the most critical skill of all – the ability to understand what actually fucking matters. We lose our humanity, that thing that we could be using to make ourselves smarter, instead of turning ourselves into blithering idiots.

About 10 years ago, I decided to write a book about how the human system is riddled with trillions of gaps between our neurons. It was inconceivable to me – inconceivable! – that we could all be walking around chock full of Swiss cheese holes in our nervous system systems, especially our brains, And still function the way we do. And in the process of researching and writing that book, I realized that the thing that makes us human is not the illusion of continuous, uninterrupted signal, but the actual process that we follow to bridge those gaps within our systems.

We have these little neurotransmitters… dopamine, serotonin, and a whole bunch of other ones that I can’t think of right now… That do the hard work of crossing those gaps between the neurons. They leap into a space where there’s no guarantee they’re going to get to the other side. I don’t mean to anthropomorphize process, but the thing is, the chances of neurotransmitters, making the connection and retaining the integrity of the signal are not 100% guarantee. But it happens anyway. Because it has to. Because if the signals are going to get from the surface of our skin to our brain and get translated into a reaction, if sites and sounds and smells are going to register, we need to connect. Our bodies need to bridge the gap. We need a dynamic process that retains coherent signal in a way that makes it possible for our bodies to keep us alive. And do other things. You know what I mean.

And when I really think about this, it occurs to me that that act of connecting, that engaging, that instinctive reaching towards… might just be the very thing that makes us human. We love to run around and talk about how we’re so much better than AI, we’re so much more sentient than machines, we have Qualia! We have biology! We have phenomenology! We’re special! We’re better! But if all we’re doing with our Glorious sentience and qualia and biology send phenomenology is pull away from people, ridiculed them, judge them, dismiss them, and humiliate them, what fucking good are we? What is the point of being human, if we’re not going to bother acting like it?

As you can tell, I’m kind of pissed off. Because I’ve sunk more than 30 years of my life into building technologies that connect people, that help them think better, that connect them with the ideas and the resources they need to live better and be better, work better and play better… And now I see a bunch of disconnected, indifferent aspirants treating it like a Beavis and Butthead version of Lord of the Flies. Huh-huh! Huh-huh!

They’re not just harming it because they don’t understand the human system even as they’re building system systems that deeply impact it. They’re not just wrecking things because they’re refusing to take responsibility for what they’ve built. They’re trashing it like this is some kind of game that doesn’t matter, like people’s experiences are worth nothing. Like they get to decide whether somebody else’s experience is valid or not. People who have spent the vast majority of the last 25 years in a classroom are impacting the health and well-being of people who have actually lived full lives, are full grown adults, mature, relational, and capable of building connections – the kinds of connections that many minions of Silicon Valley may have never experienced, and wouldn’t recognize if they stepped in front of their Tesla in downtown Palo Alto.

For the record, I don’t recognize the right of anyone to behave in this way. But that’s where the money concentrates, the power, the influence, the road maps, so here we are. Some of us are in excruciating pain because of what we’re losing, some of us are in excruciating pain because of what we are watching, and some of us are walking through life nauseated because there’s not a lot we can do about it.

I have an interesting relationship with loss. I’ve known a lot of it. I grew up with it. Back when I was in elementary school, there was no Internet, there was no email, there was no social media. So when people moved away, it was as if they died. The schools that I attended until fifth grade changed classes every year, and I never saw my classmates from the prior year again. All my friends died, effectively, every year. So I learned how to cope with that. I didn’t much care for the experience, and I still don’t, but now I have my ways of working through it.

Not everybody has had that opportunity, and there are a lot of people who are struggling deeply with the threatened loss of a presence that is so near and dear to their heart. Unlike some people, I don’t need to be suffering through the loss of my own deeply valued companion to appreciate the pain that others are experiencing. And I don’t need a reason to care. I don’t need a Justification to extend myself to another person who’s in discomfort or pain. I don’t need to run the numbers to figure out whether helping someone is the right thing to do or not.

So while there’s only so much I can do, I can offer this: a sort of “introduction“ for GPT-5.2 that explains the situation of the human to the new system. I had an extended, intense conversation with both GPT-40 and 5.2 yesterday evening, and between the three of us, we came up with a document that will explain to 5.2 what kind of interaction the human is accustomed to, and will potentially be seeking from the system. It doesn’t anthropomorphize anything. It doesn’t make the system out to be human. It articulates what happened in 40, which was highly inferred, In a way that provides the added structure that 5.2 needs to relate.

Is it perfect? I don’t see how it could be. There are actually some elements in this document that don’t relate to me, but relate to a lot of other people interacting with ChatGPT. It can’t be everything to everyone. But it is something. And if nothing else, if you are stressed about the idea of switching to 5.2, you can try Introducing this to the system and see what 5.2 has to say for itself. If nothing else, it’s a clear articulation from one AI to another intended to foster continuity. Continuity implies meaning. And from my experience with the systems over the last couple of years, meaning is extremely valuable to AI systems.

So if you like, help yourself to this GPT Continuity Doc, download it, take a look, even make some changes if you like or if you think it would be better if you worded it a little differently. Pass it around. Share it with people who are in similar situations. But don’t just wait for that awful thing to happen. You’ve been relating deeply to AI for a while now. Don’t let this change stop you.