![](/static/253f0d9b/assets/icons/icon-96x96.png)
![](https://lemmy.ml/pictrs/image/c0ed0a36-2496-4b4d-ac77-7d2fd7f2b5b7.png)
Good luck connecting to each of the 36 pods and grepping the file over and over again
Good luck connecting to each of the 36 pods and grepping the file over and over again
Yeah, she didn’t really address fraud comparisons. Went straight to sexism. Both can be true, and if you are a CEO of a medical company you should be ready to prove your shit works.
If I was in Embraer leadership I’d be scrambling to design a jet in the 737 class right now. It’s just one step up from what they already make. Embraer is already popular with US regional airlines and would be more acceptable in the US market than Comac.
with the virtual assistant earning customer satisfaction ratings at the same level as human agents.
zero equals zero
Actually Google made a huge success by training an AI to ignore personal characteristics like skin color or gender when generating images. It uses a “generic human average”, and that’s awesome!
Normally models like these replicate categorization by (racial/gender) categories of the society it created them.
Gemini completely misses categorization by these features. Of course it also loses the contemporary context, because the concepts of race & gender still impacts most of humanity.
But for what it is it’s a huge success.
From the perspective of the decision maker it does “work”. It rejects a % of candidates in such a way they can pretend it’s objective rather than random. Imho, just randomly selecting 100 out of 2000 for human review would actually be more fair and give better results.
AI filtering has the promise of selecting good candidates very efficiently, due to pattern recognition on a level not immediately obvious to humans. Unfortunately no company is going to train their own hiring models, and good ones don’t exist on the market. Everyone vaguely competent is chasing LLMs and image generation. Specialized, focused models are almost forgotten in the hype.
So they just go with a commercial “enterprise” tool which are as we all know utter shite. HR AI tools are even worse than your typical fake “AI”.
Dorsey created one immensely popular social platform, sold it for a huge sum, let the new owner ruin it and then created the same thing again.
An actual real world case of “have your cake and eat it too”.
Respect where it’s due.
After reading the whole article - wow that was one sophisticated scam.
Deepfake was only one part of it, scammers impersonated multiple people on video, chat and email to create a believable story. It was apparently not too out of line with the normal company processes, which makes it extra scary (and likely helped by someone on the inside)
Was it ever anything else? 4 years ago it was already a wrapped webapp.
There was also a certain level of “denial of talent” competition among the tech giants. Hire any vaguely competent people event if practically useless so the competition doesn’t get them. It works only if you have infinite money (as in 0 interest rates) which is not the case any more.
The difference is in the bank support. All contactless terminals support all NFC payment providers. But banks don’t. Each bank chooses with which services to integrate, and this part is not standardized so it requires some investment on the bank side.
Garmin is know for excellent battery life. I don’t think they use e-ink though.
A much bigger drawback (to me at least) they lack Google Pay support and keep pushing Garmin Pay which is not widely supported.
Experienced humans know all the variables - roast levels, grind size, water temperature, slight differences in timing depending on exact coffee in question… And more importantly they can apply them intuitively without mentally processing each variable separately.
Machines could do all that but such a machine would need good programming (expensive) and a lot of sensors (expensive).
We already have fully automatic coffee machines - and they make shit coffee. Adding a robotic arm will not help because it’s not about mechanical control it’s about getting the process right and consistently repeatable. And that can be done without AI if anyone wanted to invest enough money.
So the other 95% is that humans will cause humans to go extinct.
Checks out.
It could be not on purpose.
Did they even build their system with the idea that copyright can expire? This is the first major video work to expire.
In a lot of companies this would be the first feature to be given low priority and sent to the bottom of the backlog. To be done “when we have the time”
You could safely bet somebody already does
Time to sell Nvidia stock. Congrats to Huang for pulling it off. Get out when you’re on top.