Thanks! At first I was using OpenAI's deep research to just give a summary and overall score 1-10, but I realized that could not be iterative and future proof as new evidence comes to light.
So after some thought, I switched to a system of individual evidence gathering and weighting each piece of evidence. I've given the models some basic starting points for types of evidence (for instance a donation has a default weight of 8/10), but have given the models leeway to make relative judgements.
After all evidence is collected, the weights and confidence that the evidence is accurate (usually very high) are put into a formula to derive a final score. No recency bias. The nitty gritty:
-Each row contributes direction × weight × confidence × status_factor, where disputed is cut in half and there is no recency decay.
-All signed contributions are summed into S, and total support mass goes into M.
Final score is 50 + 50 * (S / (M + 4)), clamped to 0-100.
-That +4 prior mass keeps thin but unanimous evidence from producing extreme scores too easily.
-Neutral evidence (direction = 0) doesn’t push the score up or down, but it does increase M, which pulls the result back toward 50.
As for the ladder - I think that is a good idea, but in a controlled manner because of the token cost and potential for abuse.
Maybe you genuinely have no use case for ChatGPT. Maybe you just haven't been creative enough to figure out how to use the tech as it currently is. Forget AGI. What it is capable of right now, for me, and countless others in countless fields of work, already saves more time per day than everything else combined. There's certainly nothing more important than my time. That's a pretty powerful product.
Perhaps people can feel like they have special talents or powers in one field while still feeling apprehensive and anxious when it comes to other facets of life ie. I don't know that success in business equates to confidence on the dance floor.
I'll add to your anecdotal evidence by saying I am a vegetarian and I am interested in this. So somebody does really want this, just not people "around" you.
-Chiefs win the Super Bowl
-Bruins win the Stanley Cup
-Nets win an NBA title
-Yankees win World Series
-People continue to spell the opposite of win, incorrectly, en masse, by “sounding it out”
Generally...he'll have to lower the price on the remaining 3 shares to sell them it looks like as there was only demand for 7 items @ $10. Or he can simply wait until someone values them at $10. Conversely, if someone is willing to buy more than is available, people will probably make more available, just at a higher price.
So after some thought, I switched to a system of individual evidence gathering and weighting each piece of evidence. I've given the models some basic starting points for types of evidence (for instance a donation has a default weight of 8/10), but have given the models leeway to make relative judgements.
After all evidence is collected, the weights and confidence that the evidence is accurate (usually very high) are put into a formula to derive a final score. No recency bias. The nitty gritty:
-Each row contributes direction × weight × confidence × status_factor, where disputed is cut in half and there is no recency decay.
-All signed contributions are summed into S, and total support mass goes into M. Final score is 50 + 50 * (S / (M + 4)), clamped to 0-100.
-That +4 prior mass keeps thin but unanimous evidence from producing extreme scores too easily.
-Neutral evidence (direction = 0) doesn’t push the score up or down, but it does increase M, which pulls the result back toward 50.
As for the ladder - I think that is a good idea, but in a controlled manner because of the token cost and potential for abuse.