huginn

Quote from abstract:

We find that the MTEs are biased, significantly favoring White-associated names in 85.1% of cases and female-associated names in only 11.1% of cases, with a minority of cases showing no statistically significant differences. Further analyses show that Black males are disadvantaged in up to 100% of cases, replicating real-world patterns of bias in employment settings, and validate three hypotheses of intersectionality. We also find an impact of document length as well as the corpus frequency of names in the selection of resumes.

Pretty damning... And not surprising.

Oh my God an image macro posted to the political memes board doesn't have 60 pages of nuanced footnotes explaining how if you're in Nebraska then your voting system works differently I'm going to literally shit my own pants out of rage

Cause and effect reversed there buddy

New devs: "wow a new sdk version"

Old devs: "Ah fuck I've gotta change how we request permissions again"

You might not notice it but us Android devs do.

It is my understanding that the fundamental architecture (the general purpose transformer) is identical between the "AI" used by Black Rock and by OpenAI

If you have some evidence to the contrary I'd always appreciate the chance to learn.

But the transformer based architecture is fundamentally flawed: it will always hallucinate.

The "AI" being used by big corporations is still fundamentally an LLM and has all the flaws of an LLM. It's not a hot wheels car vs a tank, it's a hot wheels car vs a $2 billion RC car

I'm having an aneurysm trying to understand what's being said here, can anyone explain?

Good to know - I didn't opt for the 1k range personally, I sprang for a Priority Current with Enviolo hub. Love that bike - commute 9 miles to work on it.

(it's not his personal device - it's a work owned device)

Zoom ai transcriptions also make things up.

That's the point. They're hallucination engines. They pattern match and fill holes by design. It doesn't matter if the match isn't perfect, it will patch it over with nonsense instead.

I dug into this more because I was under the impression it was the case that most states did not: here's what I found

https://felonvoting.procon.org/state-felon-voting-laws/

23 states let you vote after prison. 9 fully stop you off your rights. DC & 2 State let you vote from prison. The others are some level of parole or parole and probation.

Which means overall 9/51 are full loss: everywhere else you can theoretically get it back.