Nov 29, 2022·edited Nov 29, 2022Liked by Alberto Romero
Thanks for this informative article, new subscriber here, looks good so far.
Your reasonable complaint seems to be that Galactica does not work as well as it's creators claim. Who knew that could happen? :-)
Seriously, I have a different concern. What if Galactica did work very well? What if Galactica succeeded at it's intended goal of further accelerating the knowledge explosion?
I'd be curious to learn how Meta developers might address this question. What do they see to be the logical outcome of providing human beings with ever more, ever larger powers, at an ever accelerating rate?
Who are these human beings? Hmm, let's see, I remember now, human beings are the species with thousands of massive hydrogen bombs aimed down their own throats, an ever present existential threat that even their most highly educated members rarely find interesting enough to discuss, even in presidential campaigns when they are selecting a single human being to have sole authority over the use of these civilization ending weapons.
If I had a loaded gun in mouth which I found too boring to discuss, would you consider me rational? Would you respond to my situation by giving me ever more, ever larger powers, at an ever accelerating rate?
What if Galactica actually worked? That's what worries me. I was relieved to learn that, so far at least, it seems not to.
Thanks Phil! (I've seen all your other comments, thanks for the overwhelming interest in my articles! I'll try to answer everything)
"What if Galactica did work very well?" This is an intriguing thought experiment (I don't see Galactica succeeding, but I can imagine future AIs doing so). Maybe I could explore those types of counter-intuitive thought experiments in future articles...
The principle involved here seems relevant beyond Galactica. What if AI in general succeeds in further accelerating the knowledge explosion? It seems that technologists in many fields simply assume that's a good thing, without really giving it much thought. Perhaps we should be giving this question more thought? If, on balance, further accelerating the knowledge explosion is a net negative, that would seem to put AI in an entirely different light.
I've not yet found anyone writing on this. Should you know of such writers, perhaps you could introduce us to them?
On a completely unrelated matter -- without doubt, Defense Departments around the world are spending tons on military applications of large language models. Is there anything to be found here? I can't think of anything but I know zero about the field. Speculations, anyone??
"Galactica’s problem goes deeper because sounding “right” and “authoritative” would make anyone who doesn’t have prior formation on a topic believe with illusory certainty that the newly acquired (flawed) knowledge is true." - this can easily be applied to teachers for example :(
AI like the internet is a magnifier of Humans logic & behaviour, and there will always be place to subjective POVs.
It's not the AI fault or the fault of the team behind it, that US take for granted their outputs. It's on US to be self-aware and open minded (like when reading books).
Perhaps a simpler speculation might be what is going to happen to the enormous field of virtual
sex. There are maybe thousands of young men and women out there selling pictures or videos of themselves or real time video displays. It certainly looks as though this field will be in a serious situation by this time next year. And probably before. There might be a business model left in selling prompts, but it will be much smaller. What are the implications? Will physical sex working experience a significant labor surge??
Thanks for this informative article, new subscriber here, looks good so far.
Your reasonable complaint seems to be that Galactica does not work as well as it's creators claim. Who knew that could happen? :-)
Seriously, I have a different concern. What if Galactica did work very well? What if Galactica succeeded at it's intended goal of further accelerating the knowledge explosion?
I'd be curious to learn how Meta developers might address this question. What do they see to be the logical outcome of providing human beings with ever more, ever larger powers, at an ever accelerating rate?
Who are these human beings? Hmm, let's see, I remember now, human beings are the species with thousands of massive hydrogen bombs aimed down their own throats, an ever present existential threat that even their most highly educated members rarely find interesting enough to discuss, even in presidential campaigns when they are selecting a single human being to have sole authority over the use of these civilization ending weapons.
If I had a loaded gun in mouth which I found too boring to discuss, would you consider me rational? Would you respond to my situation by giving me ever more, ever larger powers, at an ever accelerating rate?
What if Galactica actually worked? That's what worries me. I was relieved to learn that, so far at least, it seems not to.
Thanks Phil! (I've seen all your other comments, thanks for the overwhelming interest in my articles! I'll try to answer everything)
"What if Galactica did work very well?" This is an intriguing thought experiment (I don't see Galactica succeeding, but I can imagine future AIs doing so). Maybe I could explore those types of counter-intuitive thought experiments in future articles...
Hi Alberto, thanks for your reply, and your blog.
The principle involved here seems relevant beyond Galactica. What if AI in general succeeds in further accelerating the knowledge explosion? It seems that technologists in many fields simply assume that's a good thing, without really giving it much thought. Perhaps we should be giving this question more thought? If, on balance, further accelerating the knowledge explosion is a net negative, that would seem to put AI in an entirely different light.
I've not yet found anyone writing on this. Should you know of such writers, perhaps you could introduce us to them?
On a completely unrelated matter -- without doubt, Defense Departments around the world are spending tons on military applications of large language models. Is there anything to be found here? I can't think of anything but I know zero about the field. Speculations, anyone??
I don't know either, but it's an interesting topic to research about...
Love, love, love this image of the galaxy in a book.
"Galactica’s problem goes deeper because sounding “right” and “authoritative” would make anyone who doesn’t have prior formation on a topic believe with illusory certainty that the newly acquired (flawed) knowledge is true." - this can easily be applied to teachers for example :(
AI like the internet is a magnifier of Humans logic & behaviour, and there will always be place to subjective POVs.
It's not the AI fault or the fault of the team behind it, that US take for granted their outputs. It's on US to be self-aware and open minded (like when reading books).
Loved the essay, thx! :)
Perhaps a simpler speculation might be what is going to happen to the enormous field of virtual
sex. There are maybe thousands of young men and women out there selling pictures or videos of themselves or real time video displays. It certainly looks as though this field will be in a serious situation by this time next year. And probably before. There might be a business model left in selling prompts, but it will be much smaller. What are the implications? Will physical sex working experience a significant labor surge??