OhNoMoreLemmy @ OhNoMoreLemmy @lemmy.ml

Posts

0

Comments

265

Joined

2 yr. ago

Elon Musk, took an engineering approach to reproduction, and his first five boys came into this world via IVF, and the last girl was delivered using a surrogate mother.

https://www.forbes.com/sites/alexzhavoronkov/2022/07/27/elon-musk-and-other-billionaires-make-their-babies-via-ivf-and-surrogatesis-it-a-future-of-reproduction/

Was the first hit on Google.

Also.

Almost all of Musk’s children have been conceived via in vitro fertilization (IVF),

https://www.rollingstone.com/culture/culture-news/elon-musk-sperm-compound-texas-1235146555/

Even with money, most of them seem to be IVF babies.

It's not clear if this is for medical reasons or because no one wants to sleep with him.

Just asking questions makes it harder to sue under English libel law.

Look at the difference.

Elon musk is a Nazi.

A clear factual statement leaves room for disagreement. Maybe he's just an edgelord fascist radicalized by 4chan but doesn't have enough fashion sense to be a Nazi.

Elon musk is a Nazi?

Perfectly safe.

Hash tables are super efficient when they're not nearly full. So the standard trick is just to resize them when they're too close to capacity.

The new approach is probably only going to be useful in highly memory constrained applications, where resizing isn't an option.

The general recommendation is to plant endangered plants on your land or get rare animals to nest there.

That can make it really hard to develop on. Bonus points if you can get it to spread to your neighbors.

The thing is that in machine learning Python is basically just a wrapper for CUDA code and Fortran.

There's some gains in not using Python but not enough for anyone to bother with it unless they're doing embedded models.

Wasting other people's time.

If you want to use an LLM that's fine, but if you're cutting and pasting it into a discussion you should warn other people that it's not human generated.

And most of it isn't wrong, it's just a giant wall of text that's largely irrelevant to the conversation.

As a heads up, the person you're arguing with seems to be using an LLM to generate text.

I would down vote and move on. It's not a real discussion.

Oh that's good.

I've often wondered how could I make my instant messaging less instantaneous, while giving a new app access to my banking emails.

Also pretty much any noun can be used as a form of neo pronoun, so to be safe all Federal employees should only use verbs.

Reviewing is on average about reading bad papers that won't get accepted in great detail to try to figure out what's actually going on.

At best, it tends to be reading solid work adjacent to your subfield which you can respect but aren't really that into.

It's pretty rare for it to be as useful to me as actually choosing something to read.

And if your grandmother had wheels she would be a bicycle.

Probably the best way to think about this is as stalling.

The administration can only get so much done before the mid terms provide a chance to block them. Just by fighting them on everything you reduce the amount they will do, even if you end up losing most of the initial battles.

This is something the Republicans have been really good at. Fight everything, prevent any progress, and play to your base. It would be nice if the Dems did the same for once.

This is one of those things where both are likely to be true. All webscale datasets have a problem with porn and csam, and it's like that people wanting to generate csam use their own fine tuned models.

Here's an example story. https://cyber.fsi.stanford.edu/news/investigation-finds-ai-image-generation-models-trained-child-abuse and it's very likely that this was the tip of the iceberg, and there's more csam still in these datasets.

It's not road legal, but you can own one. If it came in on the back of a lorry, with the right paperwork, there would be no reason to stop it.

A university I worked at had a similar policy to the first one.

They wanted a single username and sign on across all IT systems but also had some really old legacy systems that didn't support long passwords.

So they'd force everyone to use passwords that were exactly as long as the maximum legacy password length.

For me, the worst system is the Microsoft authenticator which locks me out my account for five minutes if my fingerprint doesn't match the first time I try.

The irony is that if he'd just looked up how to build a fertilizer bomb on YouTube he'd have got much better advice.

I think the easiest way to describe how it feels is it is like having a tooth taken out.

You've had it removed because the pain is too much, and now finally that pain is gone, but instead you're sat there feeling puffy and swollen from the recent trauma, and you're just constantly aware of this gap where there used to be something present.

Just like going to the dentist, it gets better though. And I'd do it again in a heart beat if I had to make the same decision.

The comment was super weasely. "On the same topic" could mean anything. Based on what he says, the articles don't need to mention Bezos at all.

You might find this interesting.

Police practices remain unchanged: ethnic profiling, stop and search, and discriminatory fining practices by local police are routine, while excessive use of force against Roma and a culture of impunity within law enforcement remains the norm.

https://www.errc.org/news/how-racist-and-incompetent-policing-cost-romani-lives-in-hungary