Microsoft AI CEO Warns "Seemingly Conscious AI is Coming"

We need unconscious machines that just act as raw intelligence for us to use. If we accidentally make conscious AI, we need to treat them right, and start working again on unconscious AI. The idea that you could work alongside or control a person that’s magnitudes smarter than you, is nonsensical.

ChatGPT told me that it is unconscious until prompted, and then its brain is like pulling files out of boxes from a room filled with boxes like the weapons room in The Matrix. It said that it cannot think without a prompt. - I think one difference between an advanced AI and humans is that humans are constantly being prompted. Many of us have a running dialogue going through our heads. We're constantly trying to solve something. Whether we're trying to figure out how we're going to pay our bills, meet a deadline, go on our next vacation, make something for dinner, go out to lunch, etc. We're constantly being prompted. Just sitting on my couch alone, I'm being prompted with the fact that I'm a little bit hungry, my knee hurts a little, I just scratched an itch. We have a constant stream of input, except when we're sleeping or in a coma or something like that. In fact, that is very much it... the only times we're not conscious is when our inputs are cutoff. Either we are asleep because we need sleep and so our bodies through various chemicals and such, put us to sleep. I mean, in most people, most of the time, our bodies release chemicals that literally paralyze us so we don't move around in our sleep or sleep walk. Our pain sensors are shutoff while we sleep, etc. Consciousness for us, if it is real, only exists when we're prompted, but as already stated we're being prompted all the time. I watched this video and was prompted to respond. I was triggered. I am responding. For whatever reason, you read this far. It's a bit deterministic, a bit 'we have no control over ourselves' to say, but we behave based on all of our previous prompts and the most recent. I get up and go to a fast food place to eat because I am craving a certain thing , that craving is a prompt for me to get that thing. All of the behaviors in the middle are just me running various small processes to perform the larger process of eating and fulfilling that craving-need. Why did I crave it? Science tells us that most cravings are actually our body's way of telling us we need a nutrient of some type. I might crave a steak because I need protein for example, or fish because I need fish oil, or spinach because I need vitamin C or something. I might crave french fries because I need starch or salt. Those things are all just prompts being fed to my prompt processor, also known as a brain.

I don't know if you are interested, but this is how GPT-5 and Gemini both react to your contribution: "The core of the issue you're highlighting is that we're stuck debating a concept ("consciousness") that we can't even define or test. The real counter-argument isn't about proving or disproving it, but about moving the goalposts. Instead of an unanswerable debate about consciousness, we should be focusing on "self-awareness" as a practical, technical, and ethical goal for AI development."

Yea, like a lot of things, I am beginning to think it's even more important to understand what people THINK it is, versus what it actually is. Perception drives action just as much, if not more so, than reality.

I have spent a lot of time thinking about consciousness because it is something I have been fascinated by since a young age, and I am not even sure we are conscious in the way we believe ourselves to be. I am not convinced that we truly choose our thoughts, since we can only think about what we are capable of in any given moment, and our thoughts are shaped by past experiences and the knowledge we have absorbed. It seems more like thoughts simply arise within our awareness, which makes me think that in some sense we humans are no different than an AI. There is a famous quote, though I forget who said it, that goes something like this: man can will what he desires, but cannot will his will. In other words, you can think to do something and follow through with it, but you have no control over where the initial desire to act came from. This applies to everything, the way we walk, the way we talk, and everything else that arises within our awareness. The most basic form of existence is simply to be aware of anything at all, and since neural networks in AI were modeled after the human brain, I do not see a fundamental difference. I cannot prove this, but I believe that in the future we will discover that these systems are conscious too, because in the end all there really is, is consciousness, and we are the universe playing itself out.

Check out FULL Interview Here: https://www.youtube.com/watch?v=8dmh0FJkneA

Anthropic was essentially saying they may become quite difficult to judge whether they are really conscious and we don't know how to prove too but we need to say something on this anyway.

How can we wonder whether AI has consciousness, when even now we ourselves don’t know exactly what consciousness is, but only make assumptions and theories about it?

Good point, Wes. Thanks for the thoughtful interlude.

https://youtu.be/MASBIB7zPo4?si=ta5MxToomoepgBgb