sndrtj

joined 1 year ago
[–] [email protected] 12 points 1 year ago

Good riddance.

[–] [email protected] 13 points 1 year ago (4 children)

Chatgpt flat out hallucinates quite frequently in my experience. It never says "I don't know / that is impossible / no one knows" to queries that simply don't have an answer. Instead, it opts to give a plausible-sounding but completely made-up answer.

A good AI system wouldn't do this. It would be honest, and give no results when the information simply doesn't exist. However, that is quite hard to do for LLMs as they are essentially glorified next-word predictors. The cost metric isn't on accuracy of information, it's on plausible-sounding conversation.

[–] [email protected] 3 points 1 year ago

Statistics would indicate that that is a plausible scenario.

In addition, a uniparental disomy can occur as well. Here, the X chromosome was duplicated in the egg cell. So the exact same X chromosome is inherited twice.This is an error in meiosis. This could occur in XXX (with the third X from the father's side), XXY, or even XX. That latter one would be rare, for a uniparental disomy on X without a third sex chromosome would mean both egg and sperm cell had an error during meiosis.

You could also see a single X (Turner Syndrome) as a 100% dominant X-chromosome. But that may be semantics.

[–] [email protected] 6 points 1 year ago (2 children)

X-inactivation is a little bit more complicated than that. While the process of X-inactivation initiation is random, once a cell has settled on one chromosome, all its daughter cells will silence the same chromosome. The initial process happens in the early embryo, so large patches of the body have the same X chromosome silenced.

This pattern is visible in some animals. E.g. a tortoise cat's pattern arises due to the hair color gene existing on the X chromosome. Consequently, male tortoise cats are rare (XXY, XXXY etc only)

[–] [email protected] 3 points 1 year ago

Also certain fields (cough cough medicine) needs to consider more than just the p value. With any large sample size you're almost guaranteed to find a "significant" result in some test, but the effect sizes are often so tiny to be basically meaningless.

[–] [email protected] 1 points 1 year ago

That's the exact problem OP is referring to.

[–] [email protected] 5 points 1 year ago

I'm finding it useful for detecting / correcting really simple mistakes, syntax errors and stuff like that.

But I'm finding it mostly useless for anything more complicated.

[–] [email protected] 6 points 1 year ago

And the dystopia continues....

[–] [email protected] 11 points 1 year ago

This is very much the default in the Netherlands. Yes theft happens, but your license plate will be clearly visibly on CCTV meaning you will get a visit by police soon after.

[–] [email protected] 1 points 1 year ago

Would this plan remove all ads? Only then I'd consider it.

[–] [email protected] 10 points 1 year ago

No NDA, completely different sector. Stated reason was the same as in the article: "we need your full undivided attention", followed by some bullshit that they were "concerned" I would be overworking myself. Maybe they should have reduced the work load at work if they were truly concerned, as I was pulling 60+ hour work weeks at the time.

[–] [email protected] 4 points 1 year ago (1 children)

You never talk about what you did in the weekend over the water cooler?

Also: she, not he.

view more: ‹ prev next ›