Video discussion of this event by Steve Shives (known for his star trek videos but also does politics) https://m.youtube.com/watch?v=6aMQAv-JYpk
Video discussion of this event by Steve Shives (known for his star trek videos but also does politics) https://m.youtube.com/watch?v=6aMQAv-JYpk
One of my top 5 books. It’s also free to read online. https://www.rifters.com/real/Blindsight.htm
It in no way supports that LLMs can be sentient. And despite the arguments in the book that consciousness and awareness can be missing in an advanced species capable of space travel, I do not actually believe that’s true. But I enjoy the argument and speculation.
The book is highly researched and even contains a reference list of legit research articles. However it is a book of fiction and the writer took artistic liberties when needed to make an interesting story over facts.
For instance. A brain cannot contain two or more personalities because a personality is a full brain deal.
But it’s an interesting argument about cultural designations of what counts as mental illness.
Also the reason I do not think a space traveling species can exist without consciousness.
Because. Motivation.
It’s that simple.
An organism can be shaped behaviorally by the environment. That’s part of evolution. And this shaping can be unconscious.
But at a point, creative construction and ambition to exceed ones given optimal environment for a less optimal one (space) must be an intentional effort.
The scientific research and experimentation required to build complex machines requires a thinking and understanding mind. Because it requires critical thinking.
Critical thinking and creativity is a characteristic that requires a sense of self.
Even in our own history we see that it takes a specific type of person to pursue scholarly work. People who are less conformist are generally more capable of new inventions, research, and challenging acceptable beliefs of the mass. We never see the most rule following conformist being these people.
If everyone was like that, we wouldn’t survive. So diversity of mental proclivities within a species is necessary for advancement. Otherwise optimal survival would be met and stagnate.
Think of the horseshoe crab as an example.
Furthermore , I am a researcher in perception. And the field of perception is often referenced for the exploration of what is consciousness.
There are many definitions. But the sense of self is one. And a popular one.
Higher complex perception creates a sense of self.
It’s a product of the system.
The book does discuss this a bit.
I need to know my body and my actions are not the same as you. That you stand there and I stand over here.
I can perform an action and you can perform a different one that is unknown to me and not within my control.
This understanding of separateness. Of “,this is what I’m experiencing and where I am (spatially)” is something that would always emerge from higher perception. Such as that in most animals.
Maybe not in plants, fungi, bacteria, single cell microbes, etc.
But there are arguments and evidence for some of those examples as well.
As a final point. (I doubt anyone read all that).
Most people who think a probability model (current AI) is capable of consciousness usually have an incredibly simplified view of how the brain processes information.
They follow old school “behaviorist” perspectives. Or “the black box” perspective on brain functioning.
But a neuroscientist will tell you it’s not simple at all. It’s not info in, info out.
The system is changed, biologically, by the input.
The same input given twice will result in a different output the 2nd time.
And the 3rd. And how frequently the input is given or it’s temporal relation to other stimuli will also change its output.
This is because the organic brain learns. And this learning is a biological change in the actual neural structures (connections) and neurons firing potential. Every single moment the brain is physically , biologically, changing.
Computations in the brain don’t use actual math. It’s all estimates (heuristics). And these are not well understood how these computations are made. They don’t work as predicted.
There are always too many factors.
Individual motivations, including personality traits are also a factor in how the information is processed.
https://en.wikipedia.org/wiki/Hard_problem_of_consciousness
https://en.wikipedia.org/wiki/Need_for_cognition
https://en.wikipedia.org/wiki/Gray's_biopsychological_theory_of_personality
https://en.wikipedia.org/wiki/Binding_problem
https://en.wikipedia.org/wiki/Neural_coding
https://en.wikipedia.org/wiki/Hebbian_theory
It’s interesting that you point to https://en.wikipedia.org/wiki/Hard_problem_of_consciousness when the term was coined by David Chalmers, who published Could a Large Language Model be Conscious?. From the abstract:
So are we all just arguing about how likely it is, or are you arguing that current AI systems are definitely not conscious? If the latter, what do you think about the not-too-distant future ones?
I thought online learning was possible with current LLMs, just not worth the cost. I mean, you can at least fine tune offline based on previous outputs and feedback, e.g. RLHF. I feel like maybe neither should count, but can’t say why exactly. Not many end users bother with fine tuning anymore because there are usually more effective alternatives like RAG.
What do you think about agentic systems, i.e. running an LLM in a loop with a scratchpad and tools? They just write their “memories” into text files, but if you consider those text files part of the system, then the input does technically change the system. Of course, you could argue that doesn’t count because it’s no different to changing the input. So to count, it would have to store neuralese or a LoRA or something?
Agenic systems are definitely more sophisticated but still just directed programming.
Humans do not learn like machines learn.
I’ve already explained that the exact same input , put in twice into a human will not result in the same exact output.
But it would for a model where nothing has changed.
I also gave links to the binding problem and biopsychology of personality and how traits change how information is processed in humans.
I didn’t even go into neural noise or brain oscillations but that’s a whole other factor for processing information.
Computers don’t have any of that. They don’t actually perceive or understand anything.
This is why a human can produce new problem solving solutions.
Apply things unrelated to new problems.
We can think outside the box without producing more nonsense than useful outputs.
Machines produce mostly nonsense when parameters are relaxed.
Also Chalmers is saying he thinks potentially in the future. Someone could create artificial intelligence and it may , in part use LLMs.
That’s just him having an open mind about it.
I don’t share his sentiments. But I admit I’m open to changing my mind if I see some very convincing evidence that works with current knowledge and theories of neuroscience.
Because I’m not convinced that something is sentient because “it looks real”. Or “sounds like a person”.
It has to function in ways that would lead to evolution outside of human intervention and control with systems that would create sense of self and understanding.
Mathematical formulas cannot do either of those things.
A program directed by code a human put in, cannot do those things.
Its like cgi. It can look very realistic. But it’s not actually a real person.
Even when motion capture is used. It’s still just a program mimicking human movements because someone (a human) told it to.
eeeee! thank you for the link! i have too much good stuff to read now, in part thanks to you and @TargaryenTKE@lemmy.world (thank you both so much! i might disappear for a week into books but i promise to pop in for air). If i didn’t have a good choosing algorithm by now i’d be in analysis paralysis (for relatively trivial decisions: if you have multiple equally good options, flip a coin. use chwazi. roll a die. whatever works for that number. if, while doing the random number generator you find yourself hoping for a specific option, you know what you really want. if not, go with the random choice. you’re equally happy with all of them so what do you care if you randomly go with number eight? go with number eight.) One of the best problems to have (too many good choices).
Now what did you think of Echopraxia?
I’ll be honest, I’ve read Blindsight a few times and pretty sure only read echopraxia once. Like 10 years ago.
But I re-read the synopsis to refresh my memory.
I remember liking Blindsight more. But not why.
I’m also not sure which story elements I’m remembering came from which book.
Was the whole vampire arch and twist from book 1 or 2?
Can you remind me of a few specific points ? Maybe that will jog my memory. Or maybe I just need to re-read it.
So the vampire bit is used in both, in book 1 the main character journeys with one and at the end of the story starts to think Earth has been taken over by vampires due to radio transmissions he’s receiving on his long voyage back home. Book 2 begins with a prologue of a group of vampires breaking out of their holding cells, reversing the Crucifix Glitch on their captors, and then their leader eventually groups up with the main character (as well as the dad of book 1’s MC) and they all journey to the Sun (or rather, a station orbiting the sun). The second book also has that group/cult of people who are trying to make a gestalt consciousness, the Bicamurals I think they’re called.
Like ai told another commenter, I don’t like it as much as Blindsight, but I still think Echopraxia is really good, they just focus on wildly different topics.