" if we were not living in this darkest of timelines… 😎"
...where the Dominion of Canada government actively enables private corporations to pollute the commons, and doesn't event equire a tiny bit of payment towards any cleanup.
I have a note on my phone title “ask this question to Jesse at Metaviews”
Could we create a sovereign GitHub-like repository for AI agents? Where the value derived from our collective development of AI can be shared amongst the commons?
Then you drop this delightful bit of futurism. And my question feels incredibly underbaked 🙃
Not underbaked, just part of the collaborative baking process! I've been putting off writing about AI agents and the whole agentic industry that is emerging. The agentic warfare/government piece was largely theoretical, but I'm still uncertain about the current actual technology of agents. A lot of my sources are arguing it's all bullshit, so I'm waiting to learn more before writing about the details.
Relatedly, I read about a massive github dump of different prompts that AI applications are built from. It suggests that agents may not always take the form of code per se, but rather language/logic/prompting fed into the larger AI platforms/APIs like Claude and ChatGPT.
While I've written quite a bit of software in languages like assembly, C, Perl, Python, ECMASCRIPT, I find many fellow authors are too limiting on what they think of as "code".
Appreciate the effort to hold off and give us a more fulsome take. It will be interesting to see what “all bullshit” actually means.
(I have used brackets to denote where the logic ends and the madness begins: I imagine an Oz like scenario where we draw back the curtains and it is just a million monkeys at a million typewriters suspended on the event horizon of a blackhole with time dilation accounting for the seemingly rapid responses…)
This is perfect. How do we get it to happen?! And what the hell is emotion recognition tech? jesus I am way behind. But yes, all of this, so necessary to stop this global AI race to the bottom.
I've made a note to do a follow up on emotion recognition tech. It's both entirely bogus and fueling a lot of research and applications. The premise is that AI can detect our emotions from our language, behaviour, or more specifically changes in those areas. Yet researchers are consistently pointing out that it's total nonsense. Yet that hasn't stopped those with means wanting it, so it continues to be developed and deployed. 🤦♂️
Eugenics thinking has always believed that there is one true way to be human, and that outliers need to be corrected rather than the false ideology being corrected.
Facial recognition still to this day is less accurate for "racialized" faces -- that isn't a natural sciences issue, but a social sciences one, as "race" is a social construct, and it is being programmed into computers by people who have been programmed to not even realize they are programmed.
As an Autistic person, the vast majority Actually Intellegent fellow humans cannot accurately detect my emotions, so the notion that people with a Eugenics mindset can program computers to accurately detect human emotions (without declaring a large population as "not sufficiently human") is dangerous nonsense.
" if we were not living in this darkest of timelines… 😎"
...where the Dominion of Canada government actively enables private corporations to pollute the commons, and doesn't event equire a tiny bit of payment towards any cleanup.
I have a note on my phone title “ask this question to Jesse at Metaviews”
Could we create a sovereign GitHub-like repository for AI agents? Where the value derived from our collective development of AI can be shared amongst the commons?
Then you drop this delightful bit of futurism. And my question feels incredibly underbaked 🙃
Not underbaked, just part of the collaborative baking process! I've been putting off writing about AI agents and the whole agentic industry that is emerging. The agentic warfare/government piece was largely theoretical, but I'm still uncertain about the current actual technology of agents. A lot of my sources are arguing it's all bullshit, so I'm waiting to learn more before writing about the details.
Relatedly, I read about a massive github dump of different prompts that AI applications are built from. It suggests that agents may not always take the form of code per se, but rather language/logic/prompting fed into the larger AI platforms/APIs like Claude and ChatGPT.
Hmmm... And what exactly is "code per se" ...
I'm a fan of: https://lessig.org/product/code
While I've written quite a bit of software in languages like assembly, C, Perl, Python, ECMASCRIPT, I find many fellow authors are too limiting on what they think of as "code".
I think of culture and religion as "code".
Excellent point
Is DNA code? If so the bible might have been trying to tell us the truth. In the beginning was the word… code. 🤭
I'm told that the Hebrew Book of Genesis starts with "In A Beginning", rather than "THE Beginning".
So, a code fork - Once Upon A Time. 😀
Appreciate the effort to hold off and give us a more fulsome take. It will be interesting to see what “all bullshit” actually means.
(I have used brackets to denote where the logic ends and the madness begins: I imagine an Oz like scenario where we draw back the curtains and it is just a million monkeys at a million typewriters suspended on the event horizon of a blackhole with time dilation accounting for the seemingly rapid responses…)
This is perfect. How do we get it to happen?! And what the hell is emotion recognition tech? jesus I am way behind. But yes, all of this, so necessary to stop this global AI race to the bottom.
I've made a note to do a follow up on emotion recognition tech. It's both entirely bogus and fueling a lot of research and applications. The premise is that AI can detect our emotions from our language, behaviour, or more specifically changes in those areas. Yet researchers are consistently pointing out that it's total nonsense. Yet that hasn't stopped those with means wanting it, so it continues to be developed and deployed. 🤦♂️
Eugenics thinking has always believed that there is one true way to be human, and that outliers need to be corrected rather than the false ideology being corrected.
Facial recognition still to this day is less accurate for "racialized" faces -- that isn't a natural sciences issue, but a social sciences one, as "race" is a social construct, and it is being programmed into computers by people who have been programmed to not even realize they are programmed.
As an Autistic person, the vast majority Actually Intellegent fellow humans cannot accurately detect my emotions, so the notion that people with a Eugenics mindset can program computers to accurately detect human emotions (without declaring a large population as "not sufficiently human") is dangerous nonsense.