“We take very seriously the concerns that it could be used for bad purposes and have worked to avoid that,” Tang said. “We want to make sure people only use these types of technologies when they want to and that it helps them.”
This will obviously be abused decades from now. Why wouldn’t authoritarians listen in on brains that have trained transformer models based on forced textual consumption over years.
Just like we pay people money to sing for us, we'll pay people money to dream for us. The best dreamers and imagineers will sell their distilled thoughts for the rest of us to consume.
We'll have an indie scene. We'll share them on internet platforms. Remix them.
There's so much more good that will come from this than bad.
Yeah, like get this, when the world is turning to shit because of climate change and we're forced to live underground like rats, we can all collectively hallucinate what the world used to be like. It'll be grand. You'll be hallucinating, I'll be hallucinating, the AI'll be hallucinating too.
"His older work was much better," ChatGPT will proclaim. "... what older work?"
> we're forced to live underground like rats, we can all collectively hallucinate what the world used to be like
How do you know we aren't already? Someone might've invented this already and you might be living in your pod already hallucinating a full human experience from years ago. Similar concept to "the universe is a simulation" meme.
By most objective measures, this is false. More people are being lifted out of poverty than ever before. People are also less bored, more creative, and more productive.
> because of climate change
Hasn't happened yet. Barring a runaway effect, it looks like we'll have fewer species (lamentable) but more arable land. We're still in an ice age and we're nowhere close to Cretaceous climate.
We're closer to inventing transcendent intelligence than killing it off completely.
> forced to live underground like rats
This is just vivid imagination. We're not Morlocks.
This is dangerously fantastical thinking. Global heating gives us less arable land. It becomes harder to grow crops as the temperature increases. This is a much bigger negative effect than any positive effect from CO2 increases. A lot of current arable land in the tropics will become unusable.
Or are you thinking of growing crops in recently-defrosted tundra in Siberia? In which case, how are you going to get political agreement to grow stuff there, get high enough soil quality, irrigate it, and move all the farmers from, say, India to there to work it?
> Just like we pay people money to sing for us, we'll pay people money to dream for us. The best dreamers and imagineers will sell their distilled thoughts for the rest of us to consume.
[unpleasant thought]
> This is just vivid imagination.
The progression above is why I regard your vision of imagination as a consumer good so immediately off-putting. It's all glitz and glamor until you encounter an idea you don't like, and then you're immediately dismissive of it.
> It's all glitz and glamor until you encounter an idea you don't like,
It's called defeatism.
I think it's a sad way to spend our short ~60-80 years as temporarily thinking matter in a brilliant and infinite cosmos.
Downer dispositions suck the wonder and awe right out of the air.
Work. Build. Grow. If you have a complaint, throw yourself at the problem and fix it. Or maybe the problem isn't so big relative to the opportunity cost and the short scale of our lives.
In a world where almost every country has an arsenal of weapons, you are afraid of climate change? But yeah, that seems to be the new nemesis for a certain generation.
Anyway, your post reminded me of the "The Futurological Congress" by Lem.
Yes, I live on the coast, but even folks who don't live directly on a coastline can be affected negatively by failed crops and displaced people. The coast land in Guyana is currently below sea-level, and the sea level is rising further.
I've worked on a documentary about the effects of coastal erosion in Guyana, which won the 2020 Godrey Chin Heritage Journalism 1st prize.
For now the system has to be trained per person. Is the difference in brain activity across individuals significant enough to stop building a universal model?
It is not clear whether this is language dependent or not. If it is, I hope being bilingual or multilingual is a viable defense against this technology.
Obviously this is pure speculation, but I suspect it'll be a lot like cloning someone's voice. The first systems required the participant to read thousands of lines of text. Now you can do it over a phone call without the participant noticing that you're cloning their voice.
Why would bilingualism be a defense? Try throwing ChatGPT or whatever a line which combines several languages. e.g., If you ask it “¿Quel est the resultado of 四*6.1, in inglese?”, it will probably reply “4 times 6.1 equals 24.4.” I suppose AI doesn’t think in words, but expects bits of information to be related to X meaning, and from there it can predict a reply and translate that information into words we understand. I haven’t tried the above line, but if it’s not possible already there’s nothing really stopping us from making it work soon enough.
Edit: sorry, maybe I’m misunderstanding. Are you talking more about multilingual people thinking more abstractly, which would make it harder to tie information to any specific words?
If this is trained to detect language in your brain, I wonder how this fares with people like me who lack a clear "inner voice" (or "inner monologue"). I don't doubt that it can be trained on a brain to detect that brain's patterns but given how different human brains can be in many ways, I'm inclined to believe this won't be turned into a "universal brain reader" so easily.
“For a noninvasive method, this is a real leap forward compared to what’s been done before, which is typically single words or short sentences,” Huth said. “We’re getting the model to decode continuous language for extended periods of time with complicated ideas.”
Thats pretty scary. Who would work on this?
Refine this and in 10 years put it into wifi base stations that scan everything and you can read the thoughts of anyone in a room. Auto scan for "dangerous" keywords and voila: automated police state!
What people tend to forget about these dystopias is that dystopias are distributed unequally. For some people the police is already going to stop you randomly and may randomly end up killing you. Some people will already spend extensive time being searched and questioned if they want to take a flight. Some people will already be automatically rejected before anyone has actually read their qualifications or interviewed them.
The US is already a police state. The police uses military grade equipment and vehicles. They can literally get away with murder in many cases. They can rob you in open daylight, legally. They can arrest you for no crime, detain you and then dump you in the middle of nowhere. They can shoot out your eyes by directly firing rubber bullets (which are supposed to be deflected off surfaces) straight on. And if you end up in prison, you can be forced into indentured servitude.
Of course this is an exaggeration because all these things don't happen to all the people, but that's the point. Concerns about thought crime are just concerns that the apparatus could be aimed at you, too, when you've been less affected so far. The concern shouldn't be for whom it is aimed at but for the scope, scale and supposed necessity of apparatus itself. The problem isn't that we might oppress the wrong people, the problem is that we think we need oppression.
There's also an in-between status: put in the airports, police stations, tribunals, at the bank entrance, at wedding ceremonies or in churches... the fact that the scientist mentions their worries doesn't make any of that avoidable.