3 ms·

> a black man in the white house and women on the presidential ballot. The retrograde prejudices so deeply ingrained in the country’s memome that we are still i

> a black man in the white house and women on the presidential ballot. The retrograde prejudices so deeply ingrained in the country’s memome that we are still in the throes of an anaphylactic reaction to an acute melatonin allergy. This is the most obscenely wrong take I can imagine, and I'll gladly explain why I say that. Your AI-generated response plainly ignores a country in love with Michael Jordan and Whitney Houston for decades, for example, and are choosing skin deep characteristics to fit _your_ narrative, while ignoring the things voters repeatedly emphasized as damning, all of which were color blind: a complete lack of policy by the "woman on the ticket", condescension from Hillary that people not following her directives "need to get over themselves", Obama chastising black men that they must vote as he says. It's a playbook on the alienation of voters.

K0balt 2y ago
It’s interesting that you suppose my writing is AI generated? I’ve not seen AI that writes that way, but I suppose you might set it up to emulate a certain style. I’ve also been accused of using AI because I used an em dash. In that case, I guess it’s almost true, because my spell check would have put it there in place of space-dash-space. FWIW I am morally opposed to publishing AI slop where it might enter the public domain, and while I find my local models useful for organizing thoughts and prepping documents, I do not use AI to write. But thank you, I guess, for assuming a properly written long form comment must be AI? lol. On the issue of pervasive racism…. Yes, we have become comfortable with people of color entertaining us, and most have multiple ethnicities both in their background and in their social circles these days… but when Obama took office, you could hear a pin drop in many, many corners of the USA. A black man running the country, representing the nation to the world, was a bridge too far for many, many people. In my life I have a lot of intersectionality with maga and maga adjacent individuals. The vast majority are racist either openly, covertly, or accidentally, the “I have a black friend” so I can’t be racist types. Xenophobia is a huge driver of the political base… if you can’t see that, IDK what to tell you… but none of that negates the very real issues you spoke of. There were no good choices, really, only maybe less dangerous ones. Anyway, I hope your day goes well, and be careful about assuming you can dismiss everything you don’t like to read as “AI generated ” and therefore irrelevant. That’s a very dangerous and sloppy cognitive shortcut to remaining ignorant of anything outside your echo chamber. Relatable though, I think we’re all a little traumatized ATM.
- jjtheblunt 2y ago
  Well said, and apologies for assuming you’d phrased using ai. Looks like we’re describing near overlap, and i suspect i know more independently voting people than the maga sorts you have seen. The large number of independents i know voted against the race-baiting tactics shown by Obama in chastising minorities if not voting as he deemed appropriate, and similarly with Hillary. Anyway none ever mention race of candidates, but voted for the candidates not mentioning race at all. I hope you see the distinction because i found that interesting.
- JCattheATM 2y ago
  > while I find my local models useful for organizing thoughts and prepping documents What did you train them on, just your own docs?
  - K0balt 2y ago
    I typically just use agentic frameworks to make workflows using RAG. I have done some fine tuning and LORA tunes but I find that a good model with agentic prompting handles most things with a little tweaking. I’ve not done much recently as I’ve been working on hardware for a new project, but I’ve seen some really cool new frameworks out there with computer use, memory, agentic prompting, and some new forms of document digestion that I’m excited to try if I can ever finish being distracted by supply chain chaos issues.
    - JCattheATM 2y ago
      As someone who hasn't done any local AI modeling but hopes to, could you clarify or explain your first sentence a little more? Happy to do my own research if you don't feel like teaching for free though - thanks for answering :)
      - K0balt 2y ago
        Agentic workflows are where a script is used to provide guidance to the llm on the character it is playing, and therefore the context in which the interaction exists. It’s like a system prompt that changes with the application, like for example you might write a script where one agent is told that it is a phd level research assistant at MIT, The next agent is the head of the department , a critical research reviewer, reviewing the research and pointing out the flaws to the “researcher” personality, that then revises and improves the answer, in a loop, with the reviewer criticising the “research” until finally the reviewer either is satisfied or the “researcher” cannot provide any new insight. That’s a tiny oversimplification, but basically that. You combine multiple agents with different “expertise” and give them access to a vectorised database of your information, or to tools that can ingest information for llm use, and you have a couple of half drunk interns that you can task with doing stuff for you. It’s pretty cool, and sometimes also hilarious. I’ve had agents get quite testy with each other and even complain about their compatriots (they are all given names in the script). A little sexual harassment as well. (I often use uncensored models, I find that for critic / skeptic roles they are far superior). Llms ape the best and the worst of humanity, but fortunately mostly the best parts.

K0balt 2y ago

It’s interesting that you suppose my writing is AI generated? I’ve not seen AI that writes that way, but I suppose you might set it up to emulate a certain style. I’ve also been accused of using AI because I used an em dash. In that case, I guess it’s almost true, because my spell check would have put it there in place of space-dash-space. FWIW I am morally opposed to publishing AI slop where it might enter the public domain, and while I find my local models useful for organizing thoughts and prepping documents, I do not use AI to write. But thank you, I guess, for assuming a properly written long form comment must be AI? lol. On the issue of pervasive racism…. Yes, we have become comfortable with people of color entertaining us, and most have multiple ethnicities both in their background and in their social circles these days… but when Obama took office, you could hear a pin drop in many, many corners of the USA. A black man running the country, representing the nation to the world, was a bridge too far for many, many people. In my life I have a lot of intersectionality with maga and maga adjacent individuals. The vast majority are racist either openly, covertly, or accidentally, the “I have a black friend” so I can’t be racist types. Xenophobia is a huge driver of the political base… if you can’t see that, IDK what to tell you… but none of that negates the very real issues you spoke of. There were no good choices, really, only maybe less dangerous ones. Anyway, I hope your day goes well, and be careful about assuming you can dismiss everything you don’t like to read as “AI generated ” and therefore irrelevant. That’s a very dangerous and sloppy cognitive shortcut to remaining ignorant of anything outside your echo chamber. Relatable though, I think we’re all a little traumatized ATM.

jjtheblunt 2y ago
Well said, and apologies for assuming you’d phrased using ai. Looks like we’re describing near overlap, and i suspect i know more independently voting people than the maga sorts you have seen. The large number of independents i know voted against the race-baiting tactics shown by Obama in chastising minorities if not voting as he deemed appropriate, and similarly with Hillary. Anyway none ever mention race of candidates, but voted for the candidates not mentioning race at all. I hope you see the distinction because i found that interesting.
JCattheATM 2y ago
> while I find my local models useful for organizing thoughts and prepping documents What did you train them on, just your own docs?
- K0balt 2y ago
  I typically just use agentic frameworks to make workflows using RAG. I have done some fine tuning and LORA tunes but I find that a good model with agentic prompting handles most things with a little tweaking. I’ve not done much recently as I’ve been working on hardware for a new project, but I’ve seen some really cool new frameworks out there with computer use, memory, agentic prompting, and some new forms of document digestion that I’m excited to try if I can ever finish being distracted by supply chain chaos issues.
  - JCattheATM 2y ago
    As someone who hasn't done any local AI modeling but hopes to, could you clarify or explain your first sentence a little more? Happy to do my own research if you don't feel like teaching for free though - thanks for answering :)
    - K0balt 2y ago
      Agentic workflows are where a script is used to provide guidance to the llm on the character it is playing, and therefore the context in which the interaction exists. It’s like a system prompt that changes with the application, like for example you might write a script where one agent is told that it is a phd level research assistant at MIT, The next agent is the head of the department , a critical research reviewer, reviewing the research and pointing out the flaws to the “researcher” personality, that then revises and improves the answer, in a loop, with the reviewer criticising the “research” until finally the reviewer either is satisfied or the “researcher” cannot provide any new insight. That’s a tiny oversimplification, but basically that. You combine multiple agents with different “expertise” and give them access to a vectorised database of your information, or to tools that can ingest information for llm use, and you have a couple of half drunk interns that you can task with doing stuff for you. It’s pretty cool, and sometimes also hilarious. I’ve had agents get quite testy with each other and even complain about their compatriots (they are all given names in the script). A little sexual harassment as well. (I often use uncensored models, I find that for critic / skeptic roles they are far superior). Llms ape the best and the worst of humanity, but fortunately mostly the best parts.