No they’ve used them on other sets, although I think primarily on sets targeted towards girls (this frozen set for example https://www.lego.com/en-us/product/elsa-s-ice-palace-43244)
Not sure if this set would be an indication that they are looking to use the more broadly.
You don’t have to use the app, the website is fine. However, with dominos if you aren’t using one of their coupon “deals” you’re overpaying for their food.
Their whole gimic is how the “coupons” save you money, but it’s not fast food prices without them.
That’s like 50% of all tip calculations nowadays. It’s really obnoxious and feels like it’s trying to make you feel guilty for tipping an appropriate amount, but taxes and service charges aren’t part of the service.
These models are also already open sourced, you can’t stop it. Additionally all of those “requirements” would just mean that AI would only be owned by big corporations.
All I’m hearing is jailtime for Tina Belcher and her erotic friend fiction!
But seriously, i generally agree that as long as people aren’t sharing it shouldn’t be a problem. If I can picture it in my head without consequence, seems kinda silly putting that thought on paper/screen should be illegal.
So am I not supposed to watch things I like?
For the new season it took me about a year to power through it, but I doubt I’ll ever watch it again. It doesn’t mean I don’t want to watch the older seasons though.
I’m surprised it got another season. I’m not sure if there was any episode I’d call good (or even decent) from season 11. While I’ve watched the series multiple times, I don’t see myself watching this next season.
That’s a weak defense of the season. The thing that made a lot of the classic episodes classics is that they are more or less timeless. If releasing the episode a year or two later means it’s not going to land well, then they should be rethinking the writing.
A great example is the iPhone episode. Its a bad episode because the premise was hardly interesting even when it was relevant, and it hasn’t aged well at all.
I think they had that in Star Trek , called it Data.
You can also set your browser to desktop mode. For some reason viewing things from a “desktop” is okay, but God forbid you view it from your phone.
It’s possible the article is referring to how they got the training data, but as we’ve been learning from recent reporting, places like cruise were actually hiring people to monitor the “autonomous” systems.
My one complaint about my new vehicle. I can basically control everything from the steering wheel, except the A/C. For some reason that needs to be weird touch controls.
Sync thing is a great resource if you want your files to be replicated across systems. I use it for files that I might work on while in different locations and devices.
Its basically a personal onedrive or Dropbox.
So the issue is not that they don’t have diverse training data, the issue is that not all things get equal representation. So their trained model will have biases to produce a white person when you ask generically for a “person”. To prevent it from always spitting out a white person when someone prompts the model for a generic person, they inject additional words into the prompt, like “racially ambiguous”. Therefore it occasionally encourages/forces more diversity in the results. The issue is that these models are too complex for these kinds of approaches to work seamlessly.
The use of LLMs though is more similar to outsourcing than it is to a new technology. No one is talking a out changing how we program, we’re talking about changing who does the programming.
While outsourcing has had its ups and downs, I think most companies have found that skilled technical people can’t really be outsourced easily/cost money everywhere. I suspect we’ll see a similar thing here with LLMs because the core compentcy that makes programmers/engineers expensive is knowing what to do, not how to do it.
I had the same exact thought the first time I saw her. They look surprisingly similar.
I’ve used Ubooquity on and off for this. It has some nice features, but isn’t open source.
It does, but it basically reloads the app when switching. Which, if I recall correctly, means no notifications from the other account and really slow swapping between accounts. When I had to use multiple accounts I would use the app for one and a browser for the other.
So there were two different configurations the model was evaluated against.
One was where they simulated patients and had them interact in an LLM like environment. In this one the model and real physicians were evaluated using an evaluation method called OSCE.
The other was having the model and physicians diagnosis old cases pulled from journals.
While the models arguably perform better in these environments, I don’t think anyone would consider these real world situations/environments. It seems closer to “LLMs being able to pass the bar” than “LLMs have been able to pratcie law”, as we’ve seen the former, but have not seen the latter.
Additionally, Google will be on my “approach with caution” list for a while after the gemini fiasco - https://arstechnica.com/information-technology/2023/12/google-admits-it-fudged-a-gemini-ai-demo-video-which-critics-say-misled-viewers/
I think the argument is that as less people have desktops and laptops, the only people left will be more technical (otherwise they’d just use a phone or tablet). The more technical people are also likely to use Linux. So as non-technical people move to tablets and phones, technical people make up a larger share of laptop/desktop users.