Earlier quoted context omitted.
If this technology becomes accessible to courtrooms or police, they will use it. There will never be a way to encrypt thoughts.
With all the amazing advances we've seen in the recent years, I'd hope people would now stop thinking "there will never be X".
High-res image reconstruction with latent diffusion models from human brain
111–120 of 170 posts
Re: High-res image reconstruction with latent diffusion models from human brain
#112Hi everybody! We’re Joe and Ahmed and super thrilled to be launching Human Diffusion today! We’ve built an exciting new image generation system that supports economies in developing nations.
Our product leverages the latent creativity of humanity by directly fitting employees with fMRI rigs and presenting them with text inquiries through our API (JavaScript SDK available, Python soon!). Unlike competing alternatives we preserve human jobs in an era of AI supremacy.
I’d like to address rumors that our facilities amount to slaving brains to machines. This is a gross misunderstanding of the benefits we offer to our staff - they are family. Our 18 hour shifts are finely calibrated based on feedback collected through our API, and any suggestion of exploitation is flatly untrue.
Send us an email (satire@humandiffusion.com) to get early access.
Re: High-res image reconstruction with latent diffusion models from human brain
#113In 1990, there was a train-wreck Wim Wenders movie that I loved and still love called "Until the End of the World". It was about a scientist (played by Max Von Sydow) who created a machine that could record someone's dreams or visual experiences directly from the brain and play them back even to a blind person. https://youtu.be/gilzgbdk300?t=442 Anyways, the images that were depicted in this work of fiction shot in 1…
Re: High-res image reconstruction with latent diffusion models from human brain
#114Re: High-res image reconstruction with latent diffusion models from human brain
#115I immediately found the results suspect, and think I have found what is actually going on. The dataset it was trained on was 2770 images, minus 982 of those used for validation. I posit that the system did not actually read any pictures from the brains, but simply overfitted all the training images into the network itself. For example, if one looks at a picture of a teddy bear, you'd get an overfitted picture of anot…
Re: High-res image reconstruction with latent diffusion models from human brain
#116Re: High-res image reconstruction with latent diffusion models from human brain
#117Earlier quoted context omitted.
People only think in words right before they say something, so I'm not sure how big a deal this is. I guess they'd be able to predict what I'm writing half a second before I write it? Would be useful if I lost the ability to write or speak, for whatever reason.
I'm "thinking in words" this entire thread as I read it. Do some people read without hearing the words in their head?
You might classify talking to yourself as a monologue, but when most people discuss this topic, it sounds like they're describing a dialogue (i.e. one between multiple people). That seems crazy to me, because who does the other voice belong to?
If you have an inner monologue, then by definition you should be able to predict what it's going to say - because mono- means just you. Yet people talk about this experience like there is some novel conversation happening in their head.
It's frightening to think about a voice in my head that is one thought ahead of me. If you experience this, how can you possibly feel in control of your own mind? When does an inner monologue become schizophrenia?
Re: High-res image reconstruction with latent diffusion models from human brain
#118Earlier quoted context omitted.
People only think in words right before they say something, so I'm not sure how big a deal this is. I guess they'd be able to predict what I'm writing half a second before I write it? Would be useful if I lost the ability to write or speak, for whatever reason.
Citation definitely needed. I have a nearly constant internal monologue that is 100% composed of words.
Re: High-res image reconstruction with latent diffusion models from human brain
#119here we see, basically, a potential feedback loop. AI tools advance brain science -- more advanced brain science can then inform progress in AI. this is why the situation is dangerous: because people dont think about these feedback loops. people see AI and they move the goalposts and rationalize by saying that "cutting edge AI is still short of AGI so its ok." but most normal people dont think about how AI can be use…
> we need regulation
No, we don't. Regulation doesn't stop technological progress - it puts it in the hands of an elite few. And besides, there are 130+ regulatory jurisdictions. For example, the US government doesn't fund human cloning research, but that doesn't mean China won't fund it. Or perhaps you'd also like a one world government that can jail anyone doing wrongthink on their GPU?
Personally, I hope we get AGI (in the most Kurzweillian sense) as soon as possible. It will lead to a cambrian explosion of advancements across all fields of science. This is our best chance of cracking the secrets of the universe and answering fundamental questions, like whether FTL interstellar travel is truly impossible, or whether aging is really irreversible.
Imagine an intelligence unencumbered by the "technical debt" we've accrued over centuries of building our scientific model of the world. AGI could simulate infinitely many novel paths through the "tech tree" of human history, replaying our scientific discoveries and trying different assumptions. What if we had 12 fingers and mathematics started from a base-12 system? What if we could see in infrared? We would have followed entirely different scientific paths; AGI will be able to find what we missed.
Re: High-res image reconstruction with latent diffusion models from human brain
#120Earlier quoted context omitted.
I'm "thinking in words" this entire thread as I read it. Do some people read without hearing the words in their head?
Yeah, personally - and this might be a controversial opinion - I think most people who say they have an inner monologue are actually misattributing their experience of subvocalization while either reading text, or planning hypothetical conversations with other people (like a child might speak to their imaginary friend). It seems dubious to label that a monologue, because it depends on external stimuli (either the tex…
No, we aren't. Now its true that (well, for me at least) the inner monologue is the exact same experience as when reading text or when planning hypothetical conversations. Or when thinking ahead of words to write. The difference is that it is not planning a hypothetical conversation or something to write, and there is no text being read, and it happens pretty much all the time, except when I'm doing one of those other things (and sometimes as an intrusive interruptions when I am.) If you imagine there exists a common piece of mental infrastructure that is used for each of those actions, its as if it was always on doing a narration except when you are specifically, actively concentrating on using it for some other purpose.
> Hearing voices is schizophrenia.
No, hearing voices that aren't there is an auditory hallucination. Among the things it can be a symptom of is schizophrenia, but "X can be a symptom of Y" is not the same thing as "X is Y".
But an inner monologue is not an auditory hallucination. Its obviously and distinctly internal, not something that "sounds" like it is coming from outside.
> You might classify talking to yourself as a monologue
Because it literally is.
> but when most people discuss this topic, it sounds like they're describing a dialogue (i.e. one between multiple people). That seems crazy to me, because who does the other voice belong to?
I think a lot of people do what amounts to roleplaying out conversations, particular on decisions which are troublesome, with themselves; because this is similar to planning a hypothetic conversation with another person, which I gather people without inner monologues can do without outward speech, I'm not sure how connected it is to an inner monologue. From the perspective of someone with one, its a fairly easy deliberate "mode switch" where you basically decided that that is what the monologue is going to focus on.
It's also conceivable that, within plural systems, there is what amounts to an "inner dialogue" or "inner multiparty conversation". Not being a plural system, I can't comment on that and the degree to which it is perceptually different than an inner monologue.
> If you have an inner monologue, then by definition you should be able to predict what it's going to say - because mono- means just you.
That...doesn't follow. But if you have an experience of a conscious thought of what you are about just instantaneously before you say it, an inner monologue is a lot like that, but without the follow-through of speech. That is, I think you are not only wrong that it is true "by definition" that you should be able to predict an inner monologue, and that this is a false analogy to external speech, but that an inner monologue is perceptually similar to, and may well be fundamentally resusing the same infrastructure as, the natural internal "prediction" (or planning; not sure those things are, in this case, different) of outward-directed speech.