Researcher in the psychology of insight, emotion, and personality.
Founder of McClafferty Innovations and inventor of NerMI, a personal intelligence system.
This is made worse when you don’t have a second dimension to cleanly separate it visually. Looking at two overlapping bell curves with a significant mean difference would usually produce the conclusion that the groups are different. For risk profiles in medicine that wouldn’t even be questioned by most scholars. But if it’s politically inconvenient, then it’s never a group difference.
@curtdoolittle I guess it applies to both local devices running it and also that the AI is specialized/focused to you as opposed to the Internet being very much the same for everyone.
I think the sudden frenzy for data centers is the wrong direction. As AI and computer chips advance (especially NPUs), I see most of what everyday people actually use being local to their device. The actual AI people use will be simpler and more useful just because of integration.
I think Runcible will help it reach the point of accuracy and safety, with simpler models enabling this direction.
The computer-human interface is what will really make the difference. AI needs to have context relevant to everyday people (like realtime emotion and current activities) and be responsive to when we need it.
@frandalorian@MelanyBelanyBoo@Timcast It’s pretty basic infant psychological development. Attachment to the mother’s voice occurs before they ever hear her voice outside the womb.
@curtdoolittle This sounds more like a bookmarking, note taking, or delegation issue. Unless you really are doing that many separate projects at the same time.
@curtdoolittle Blank slate has been dead in psychology for a while. Especially developmental psychology. The only valid questions are which genes and how much can we influence positive development (either through accommodations/correction or interactionism).
I mean, I don’t think they’d use those words, but it’s always been about accountability and responsibility. I think that’d universal all but commies, victims, and some libertarians. The people who follow you are just more likely to refer to it as a restoration of the West or be able to articulate further.
@MikeCyples@Timcast I think most of his morals was just the fact that he was mostly funded by small time donors. The talent that separated him was the ability to fund raise without selling out. (Others have been good people entering government but don’t get far due to bad fund raising.)
It has been a while since I tried Claude but I’m always impressed by how little other models can follow a request. Gemini can come close but it still ends up making major errors that are hard to notice until you’ve moved on to make dependent code. While GPT does comprehend better, I’ve found that only codex can correctly maintain context and not break the code (forcing 5-10 more prompts just to fix it).
We don’t even have to think or propose this as male versus female thinking. All that matters is responsibility. Simply requiring the draft registry to vote or requiring that they pay any amount of taxes and aren’t on welfare will make a big difference.
I’m not sure I like the idea of “net tax payer” because plenty of irresponsible people are getting fancy high paying DEI jobs. Another metric to consider is business ownership (equity over a certain percentage) possibly requiring it to be a net tax paying business (or some how profitable to avoid shell company’s or fake nonprofits; maybe exclude companies receiving government income other than bidding based work).
From experience I can say that after elementary school (5th grade) you get little more than review to a slightly higher level. By the time you’re in high school it is either entirely review or not relevant to most work. Is really just college prep if not college (lucky if trade skills are included as classes).