> My AI is a natural extension of me. It feels like a sixth sense and another limb. I wonder about a problem, feel as though I’m literally surfing the web, see glimpses of the websites, get flashes of intuition about the problem, and ultimately derive the answer.
Help me understand... If one uses AI for anything but coding, or purely utilitarian purposes, they are psychotic?
This is not intended to be provocative. I am asking sincerely. I have a lot of AI sessions, that I value enough to have reviewed multiple times, over months. And indeed, I explored much that I never would have without the AI, and enjoyed it. Even appreciated it. The bulk of my saved material is research, but at least a gigabyte is wild speculation, philosophy, and assisted thinking.
At what point does what the user thinks is a meaningful experience become psychosis, or is it that exactly; attributing meaningfulness is the symptom?
Edit: The "no longer other" statement, I should admit, is familiar to me. There can be, on rare occasion, what I have thought of abstractly as harmonic resonance, some kind of strange emergent state between two interlocutors. Maybe that is the psychosis?
“meaning” emerges from a mental process. That mental process can consume things which were built to convey meaning (e.g. a novel) or were not (a waterfall).
Just because an AI bot says things that are meaningful, that doesn’t mean you need to assume it is intentional, sentient, or intelligent. It could vomit out a random Rorschach test and if you found meaning for yourself, you can still understand it was random.
I talk to a bot to sharpen my thinking, not to adopt its thinking.
"I feel good about the vision, because it builds towards a more human future" I wonder if the author truly believes this. I could maybe buy some sort of inevitability argument, but this current framing is very hard to take seriously.
Well, the text-to-thought direction feels less sci-fi, and combining both would amount to kind of telepathy.
But sure, you're right. Reading is not a deterministic/injective mapping of text to thoughts.
Apart from the trouble of defining what a thought is, independent from the person who thinks it.
Me thinking of grabbing a mug of coffee is a different thought from you doing the same... defining "thought" pulls in a whole jungle of epistemology (maybe I should say philosophy, more broadly, I'm not in the field).
Is there an abstract concept of "the thought of raising your left index finger"? Or "the thought of remembering severe pain"? Etc etc
next they're going to set up thought-triggered Amazon purchases and say that moving the object from the warehouse to your home constitutes telekinesis.
ever since Elon did his "promise full self-driving every year" thing, entrepreneurs been on a weird sort of fiction escalator.
thought-to-text is incredible. imagine doing something that remarkable with your life and feeling the need to exaggerate.
What is the benefit ?
Sharing thoughts to the external world would be like going to a bakery and walking out with flour and eggs.
Isn't the act of talking/writing part of the thinking ? Am I just going to share intrusive thoughts and half-baked ideas to my CIA agent now ?
2040 : I earnestly believe that presidential term limits should be abolished, and also that federal control of elections is incredibly important. Not sure where that thought came from, but I'm certain!
I would like it if my AI glasses constantly recorded and compacted my reality. Then educated me as I go That way I can avoid the info dump session that never lines up with my perceived reality.
As employees they have unlimited usage IIRC, and also these kinds of tasks can be surprisingly token efficient because a very large chunk of time is spent waiting for the calculation to finish, and AI agents love to run calculations excessively.
Not sure how valuable it would be for interrogations of state level actors? Ironically, I'd think it might even be better for your adversary if the technology does work?
We know that we can use classical techniques, (non-AI), to psychologically condition humans. In the lightest case, we do it with advertising and social media. I'm absolutely certain that state-level adversaries use even more effective techniques to train and condition agents for fieldwork.
If your adversary knows you are more likely to rely on intelligence gathered in this fashion due to its reliability, they can definitely game that. Or rather, I guess, I would definitely game it if I knew the Russians had this kind of technology. So I don't know why they wouldn't be thinking along the same lines.
Basically, make them think their new technology is working.. until it counts. If it consistently works 99 out of 99 times. They'll have no reason to believe it's not working the 100th time.
But I guess if you were interrogating, maybe a pot dealer or spouse muderer or cheating husband or something, it might be valuable.
The Conduit® Thinkerface® Pro Band will contain a small 90B dense local model that pre-screens your Conduit® NeuroTokens® stream for compliance against the Conduit® Acceptable Usage Policy applicable for your region and subscription tier.
If non-compliance is detected you'll be promptly notified via a small current applied to the Conduit® ThinkRight® Human Training Electrodes embedded within your band. Intensity and duration are dynamically adjusted until compliance is restored.
Even better, because this all happens on-device before your Conduit® NeuroTokens® are sent to the Conduit® NeuroServer®, non-compliant thoughts don't count towards your Conduit® Daily Thinking Allowance or expend credits from your Conduit® Additional Thinking Purchase Plan.
Yeah, this whole area of UX scares me. Silent speech techniques could strike a balance, but it seems obvious it wont stop there - consider what a state actor would do if they could skip interrogation altogether?
Unfortunately I think this is inevitable because as a species we perform a steady march toward collective desires for items of science fiction, so start your adversarial double-think practice now.
Is it really collective desire or are these products being pushed upon us? And why does it always have to be the dystopic "destroying human nature and society as we know it" science fiction and not the utopic "everyone lived happily ever after" science fiction? Or does it just feel that way to me?
For a utopia to exist, everything has to go right. Dystopias are almost all modelled on some existing facet of human behaviour. Naturally one state is unstable as it requires practically complete co-operation in most cases while the other one is stable as it usually exists as some sort of undesirable equilibrium between competing forces.
So is it just that these are people who haven’t personally experienced well-intentioned technology turning into a living nightmare? Watching the optimism and then collapse of Web 2.0 was a formative experience for me. I’m 38. Is this just naivete? What is this?
Inevitablism, a fundamentalist belief system that resolves every moral hazard with an appeal to the inevitable. "we can't let someone else do this first"
According to her Wikipedia page, she's ~23 and went to Harvard, so A) her life experience is limited, B) most of the experience she does have involves being on a path to elitehood, and C) she probably sees a chance at a life-changing amount of money here.
People like this are the input side of the "oops we built the Torment Nexus" pipeline.
With some technologies, the negatives are so severe it is difficult to see how they outweigh the positives. Mind-control technology is quite frankly just asking to be abused. It's the ultimate relinquishment of power - you reveal your entire hand to someone else AND let them do what they want with it. I honestly don't understand how some people are so optimistic that they think this technology won't be misused in ways that massively overshadow the good that could be done with them.
"Die Gedanken sind frei..." - can someone guess them? A truly dystopian question.
It is been 10 years ago when I looked through some research in that area. Back then it was the problem to get enough good signal. Has the situation changed radically here for non-invasive electrodes or portable detectors?
I understand that the approach would be to turn weak signal, context, and LLM based signal processing into useful human computer interaction. For everything else, I have the feeling we are still in speculative territory.
I think this is actually marginally less evil than Meta Tribe. The whole neuro space is dangerous and while people go on about “AI safety” the world will trend towards this madness.
And this will provoke “well we need to regulate AI then”. No, you need to regulate poking around in people’s brains without asking them before that becomes normal.
TIL -- dear god that's terrifying. I get that there's likely a lot of valuable and legitimate work that could come out of it, but I wouldn't trust Zuck as far as I could throw him out a window.
> It’s hard to collect enough data using invasive methods. Few people want a chip in their brain! But non-invasive methods are getting much better. The hardware is improving and getting cheaper, though I apologize for being vague about the particularities of our hardware.
Pretty sure that also involves tuning your motor control in your hand and arm. It's like thinking a download into your brain would make you a successful athlete.
I'd like to think that actually-existing-telepathy would be a more transformative technology than something that simply allows people to do their existing work more conveniently.
who holds the copyright on the transcriptions of my thoughts read by this hardware? what license to those do you impose through EULA? what privacy controls need to be invented to prevent literal thought police?
> In a couple years, the main way we'll talk with our AIs is with our thoughts. They will not be a super-smart automated intern or coworker that you desperately try to keep up with. They will be a natural, joyful extension of you.
Ah, yes, very joyful indeed.
Also, fuck those interns! They're just too smart to keep up with.
Maybe I’m just demented but i feel like the first time I interface with chatgpt via telepathy it’s going to see my thoughts when the guy in the cube behind me farts or chews loudly, and send the FBI.
> My AI is a natural extension of me. It feels like a sixth sense and another limb. I wonder about a problem, feel as though I’m literally surfing the web, see glimpses of the websites, get flashes of intuition about the problem, and ultimately derive the answer. It feels fun! My brain is like the Flash. Because of write tech, my thinking is the fastest it's ever been even when I turn my AI off.
it's in the article, around 2030 the device should be able to write and "make changes to the brain: hear audio, or tell the brain to use more calories" ...and later she envisions chatGPT becoming an extension of the brain like a sixth sense.
104 comments:
> The AI is no longer “other”
> My AI is a natural extension of me. It feels like a sixth sense and another limb. I wonder about a problem, feel as though I’m literally surfing the web, see glimpses of the websites, get flashes of intuition about the problem, and ultimately derive the answer.
We call this AI psychosis.
Ironically, or sadly (or perhaps a mixture of these+), this statement was absolutely informed (hopefully not inspired) by an LLM.
Hardly a novel concept in technology: https://www.theguardian.com/technology/2016/feb/19/iphone-ap...
Help me understand... If one uses AI for anything but coding, or purely utilitarian purposes, they are psychotic?
This is not intended to be provocative. I am asking sincerely. I have a lot of AI sessions, that I value enough to have reviewed multiple times, over months. And indeed, I explored much that I never would have without the AI, and enjoyed it. Even appreciated it. The bulk of my saved material is research, but at least a gigabyte is wild speculation, philosophy, and assisted thinking.
At what point does what the user thinks is a meaningful experience become psychosis, or is it that exactly; attributing meaningfulness is the symptom?
Edit: The "no longer other" statement, I should admit, is familiar to me. There can be, on rare occasion, what I have thought of abstractly as harmonic resonance, some kind of strange emergent state between two interlocutors. Maybe that is the psychosis?
“meaning” emerges from a mental process. That mental process can consume things which were built to convey meaning (e.g. a novel) or were not (a waterfall).
Just because an AI bot says things that are meaningful, that doesn’t mean you need to assume it is intentional, sentient, or intelligent. It could vomit out a random Rorschach test and if you found meaning for yourself, you can still understand it was random.
I talk to a bot to sharpen my thinking, not to adopt its thinking.
Agreed.
"I feel good about the vision, because it builds towards a more human future" I wonder if the author truly believes this. I could maybe buy some sort of inevitability argument, but this current framing is very hard to take seriously.
From the article.
>Thought-to-text, aka telepathy
No, that's not telepathy. I'd leave the term alone. Consider it to be reserved for... well... telepathy!
Well, the text-to-thought direction feels less sci-fi, and combining both would amount to kind of telepathy.
But sure, you're right. Reading is not a deterministic/injective mapping of text to thoughts.
Apart from the trouble of defining what a thought is, independent from the person who thinks it.
Me thinking of grabbing a mug of coffee is a different thought from you doing the same... defining "thought" pulls in a whole jungle of epistemology (maybe I should say philosophy, more broadly, I'm not in the field).
Is there an abstract concept of "the thought of raising your left index finger"? Or "the thought of remembering severe pain"? Etc etc
Next you'll want the term "Artificial Intelligence" reserved for actual artificial intelligence.
Neuralink to Neuralink users
next they're going to set up thought-triggered Amazon purchases and say that moving the object from the warehouse to your home constitutes telekinesis.
ever since Elon did his "promise full self-driving every year" thing, entrepreneurs been on a weird sort of fiction escalator.
thought-to-text is incredible. imagine doing something that remarkable with your life and feeling the need to exaggerate.
What is the benefit ? Sharing thoughts to the external world would be like going to a bakery and walking out with flour and eggs. Isn't the act of talking/writing part of the thinking ? Am I just going to share intrusive thoughts and half-baked ideas to my CIA agent now ?
I can see legit use cases where people simply can't communicate in any physical way. People with locked-in syndrome, etc.
But, obviously those only make up a small number of people.
Strapping a mind-reading machine onto some detainee is probably the more likely case.
This might be naive, but I think the idea is to replace your keyboard
I knew you were going to do that
> I want to feel what my Codex feels.
Yeesh. I guess that makes one of us. Truly one of the more horrifying short stories I’ve read.
This seems strictly worse than speech to text as input to AI?
I cannot comprehend the output from AI fast enough for the speed difference between a verbal and thought response to matter.
Not to mention the privacy nightmare of logging my raw thoughts in a corporate environment and potentially routing them to an LLM provider.
I thought this was a parody when I opened it.
I think I think I think it will be eventually.
Same. I'm still not sure it isn't.
2040 : I earnestly believe that presidential term limits should be abolished, and also that federal control of elections is incredibly important. Not sure where that thought came from, but I'm certain!
What's a president? King Trump II has been our leader as long as I can remember!
King? I thought it was Lord Trump! His AI generated personality has ruled over the earth for all time.
As though we haven't done enough already to manipulate the mind, now we can have a neural model dictate our thoughts for us.
nice double entendre on "dictate", there.
We can finally have ads in our dreams like Futurama.
Probably more useful for interrogation than coding.
Under current US law, can people entering the US be subjected to mind-reading?
No. This is the philosophical basis of the fifth amendment.
> My GPT-7 agent... I become .... annoyed at the AI-speak in its first paragraph
When we have GPT-7 it will still do AI-speak?!
I would like it if my AI glasses constantly recorded and compacted my reality. Then educated me as I go That way I can avoid the info dump session that never lines up with my perceived reality.
Me too. I have meta glasses and it caps at 3 minutes and theres no auto recording.
Id be curios to see attempts of my nonverbal thoughts in writing. It doesn’t seem possible imo.
> My GPT-7 agent has been chugging away at the new encoder I’ve been exploring for over a day now.
How much are those people making to be able to afford things like this?
As employees they have unlimited usage IIRC, and also these kinds of tasks can be surprisingly token efficient because a very large chunk of time is spent waiting for the calculation to finish, and AI agents love to run calculations excessively.
If the employer is giving this to the employee, it will be income. somewhere this will have to be defended.
Visit from CIA in 3, 2, 1...
The only thing I would actually want this for would be a dream logger. Extremely invasive and quite dangerous otherwise.
There is probably good money in upgrading the polygraph security clearance ritual for the 21st century.
It doesn’t even really need to work of course.
Not sure how valuable it would be for interrogations of state level actors? Ironically, I'd think it might even be better for your adversary if the technology does work?
We know that we can use classical techniques, (non-AI), to psychologically condition humans. In the lightest case, we do it with advertising and social media. I'm absolutely certain that state-level adversaries use even more effective techniques to train and condition agents for fieldwork.
If your adversary knows you are more likely to rely on intelligence gathered in this fashion due to its reliability, they can definitely game that. Or rather, I guess, I would definitely game it if I knew the Russians had this kind of technology. So I don't know why they wouldn't be thinking along the same lines.
Basically, make them think their new technology is working.. until it counts. If it consistently works 99 out of 99 times. They'll have no reason to believe it's not working the 100th time.
But I guess if you were interrogating, maybe a pot dealer or spouse muderer or cheating husband or something, it might be valuable.
- telepathic communication with your computer
- still annoyed by AI-speak
What a bold vision!
Hmm...how do you stop private thoughts from reaching the LLM? Suddenly you can be prosecuted for thinking the "wrong" thoughts?
That's easily solved!
The Conduit® Thinkerface® Pro Band will contain a small 90B dense local model that pre-screens your Conduit® NeuroTokens® stream for compliance against the Conduit® Acceptable Usage Policy applicable for your region and subscription tier.
If non-compliance is detected you'll be promptly notified via a small current applied to the Conduit® ThinkRight® Human Training Electrodes embedded within your band. Intensity and duration are dynamically adjusted until compliance is restored.
Even better, because this all happens on-device before your Conduit® NeuroTokens® are sent to the Conduit® NeuroServer®, non-compliant thoughts don't count towards your Conduit® Daily Thinking Allowance or expend credits from your Conduit® Additional Thinking Purchase Plan.
As someone with OCD, this tech is pretty far up there in terms of potential nightmares.
Yeah, this whole area of UX scares me. Silent speech techniques could strike a balance, but it seems obvious it wont stop there - consider what a state actor would do if they could skip interrogation altogether?
Unfortunately I think this is inevitable because as a species we perform a steady march toward collective desires for items of science fiction, so start your adversarial double-think practice now.
Is it really collective desire or are these products being pushed upon us? And why does it always have to be the dystopic "destroying human nature and society as we know it" science fiction and not the utopic "everyone lived happily ever after" science fiction? Or does it just feel that way to me?
For a utopia to exist, everything has to go right. Dystopias are almost all modelled on some existing facet of human behaviour. Naturally one state is unstable as it requires practically complete co-operation in most cases while the other one is stable as it usually exists as some sort of undesirable equilibrium between competing forces.
Just wait until advertisers can get in on your real-time mental stream of thought.
This is a Black Mirror episode coming to life.
"My AI just sent my boss a 12 paragraph screed because I thought about how much I hate him"
"In a couple years"
In a couple years?! Get over the technical hurdles, do you think PEOPLE will be ready to telepathically talk to a computer in A COUPLE OF YEARS?
"probably in our lifetime"
https://www.youtube.com/watch?v=jroQCyWwEgE
vim already exists
secret counterexample. people have been failing to learn vim for literal decades
Upon trying it for the first time more than 60% of people never exit - that's just great retention!
Every time I accidentally start vim, I just buy a new computer.
You mean 60% have still not discovered :q!
So is it just that these are people who haven’t personally experienced well-intentioned technology turning into a living nightmare? Watching the optimism and then collapse of Web 2.0 was a formative experience for me. I’m 38. Is this just naivete? What is this?
Inevitablism, a fundamentalist belief system that resolves every moral hazard with an appeal to the inevitable. "we can't let someone else do this first"
Like Edward Teller wanting to develop hydrogen bombs while Oppenheimer was trying to prevent an arms race from starting.
"I will kill you now since you're dying anyway so just hand me all your money"
It's called being insulated from the immediate consequences of their actions and hubris.
According to her Wikipedia page, she's ~23 and went to Harvard, so A) her life experience is limited, B) most of the experience she does have involves being on a path to elitehood, and C) she probably sees a chance at a life-changing amount of money here.
People like this are the input side of the "oops we built the Torment Nexus" pipeline.
> Watching the optimism and then collapse of Web 2.0
What collapse of web 2.0?
Technically the website we're on is web 2.0 because we're interacting as users.
I don't think it's naiveté on their part to want to build new things. I think you're just jaded and hyperfocusing on the negatives.
With some technologies, the negatives are so severe it is difficult to see how they outweigh the positives. Mind-control technology is quite frankly just asking to be abused. It's the ultimate relinquishment of power - you reveal your entire hand to someone else AND let them do what they want with it. I honestly don't understand how some people are so optimistic that they think this technology won't be misused in ways that massively overshadow the good that could be done with them.
I understand where you are coming from. Yet.
Neuralink exists since 10 years, is in clinical trial, has been implanted on a living human being in 2 years ago.
The lack of adults in the room to set rules around moral hazard is hardly her generation's fault.
To me, this reads as an attempt to raise easy investment money.
"Die Gedanken sind frei..." - can someone guess them? A truly dystopian question.
It is been 10 years ago when I looked through some research in that area. Back then it was the problem to get enough good signal. Has the situation changed radically here for non-invasive electrodes or portable detectors?
I understand that the approach would be to turn weak signal, context, and LLM based signal processing into useful human computer interaction. For everything else, I have the feeling we are still in speculative territory.
well she is only 23 so naivete would make sense, although it looks like she held some AI safety roles so...
In the same way that “Anthropic” apparently means misanthropic, we can probably add “AI safety” to the list of newspeak.
I think this is actually marginally less evil than Meta Tribe. The whole neuro space is dangerous and while people go on about “AI safety” the world will trend towards this madness.
And this will provoke “well we need to regulate AI then”. No, you need to regulate poking around in people’s brains without asking them before that becomes normal.
TIL -- dear god that's terrifying. I get that there's likely a lot of valuable and legitimate work that could come out of it, but I wouldn't trust Zuck as far as I could throw him out a window.
are you specifically focused on non-invasive methods/why? seems like you reach a noise barrier which limits things like this
From the article
> It’s hard to collect enough data using invasive methods. Few people want a chip in their brain! But non-invasive methods are getting much better. The hardware is improving and getting cheaper, though I apologize for being vague about the particularities of our hardware.
This is doomed to fail. Dry EEG electrodes basically only capture noise.
You know how a microphone can also serve as a speaker?
Can this technology also do text to thought?
sure would be cool to transfer memories. Then we'd be able to download how to fly a helicopter.pdf right to our brains like in the matrix.
Pretty sure that also involves tuning your motor control in your hand and arm. It's like thinking a download into your brain would make you a successful athlete.
I'd like to think that actually-existing-telepathy would be a more transformative technology than something that simply allows people to do their existing work more conveniently.
If it doesn't work try thinking in Russian
It doesn't seem to bode well that early 20-something year olds who think it would be cool if AI could read your thoughts are so involved in AI safety.
who holds the copyright on the transcriptions of my thoughts read by this hardware? what license to those do you impose through EULA? what privacy controls need to be invented to prevent literal thought police?
if you get an earworm you will have to pay royalties
> In a couple years, the main way we'll talk with our AIs is with our thoughts. They will not be a super-smart automated intern or coworker that you desperately try to keep up with. They will be a natural, joyful extension of you.
Ah, yes, very joyful indeed.
Also, fuck those interns! They're just too smart to keep up with.
Never mind "how could this go wrong". How could this go right?
Simple, think about going right
For some reason your comment reminded me of https://www.youtube.com/watch?v=ky-uzsw0kqw
Up and to the right, to be more specific. Transcend your one-dimensional thinking, Razengan.
Maybe I’m just demented but i feel like the first time I interface with chatgpt via telepathy it’s going to see my thoughts when the guy in the cube behind me farts or chews loudly, and send the FBI.
What a terrible idea.
I can't wait for this tech to be incorporated into corporate productivity trackers that seem to be getting more popular.
> My AI is a natural extension of me. It feels like a sixth sense and another limb. I wonder about a problem, feel as though I’m literally surfing the web, see glimpses of the websites, get flashes of intuition about the problem, and ultimately derive the answer. It feels fun! My brain is like the Flash. Because of write tech, my thinking is the fastest it's ever been even when I turn my AI off.
I read "race conditions" all over this vision.
My thoughts being beamed to a frontier model where they are monetizable (and used for training) is terrifying.
They will also want to beam back the response straight back into your brain one day.
it's in the article, around 2030 the device should be able to write and "make changes to the brain: hear audio, or tell the brain to use more calories" ...and later she envisions chatGPT becoming an extension of the brain like a sixth sense.
The future is Borg.
I am locutus.
No worries, I'm sure there won't be Flock cameras reading your thought crimes as you drive/walk by.
Translation:
"I am a former-OpenAI employee and looking for investors to fund my AI startup."
I hope you are unsuccessful in your authoritarian, dystopian enterprise.
> non-invasive
Dead in the water.
It's like an omlette. Can't make a neural interface worth a damn without cracking open a few skulls.
> Can't make a neural interface worth a damn without cracking open a few skulls.
I immediately thought, ironically, your neurology wrote that with a computer and HN humans' neurologies read it with computers.
exciting! you could imagine brain transcriptions being very high signal RLHF data that could help align models more deeply to human preference.
Why I'm Leaving OpenAI to build the Torment Nexus
I don't think I've ever been so viscerally revolted by the vision of a company.
If this dystopian future is ever to come, I think those dates need a +100.
Thanks, I hate this.