I don't see like that, for me it's like: here's an approximate answer. Is that sufficient or do you need a 100% accurate number? I can get this info for you, but I will have to stop everything I'm doing right now to figure it out or do it after I'm done with other priorities.
Sometimes the person asking just needs an estimate, a ballpark number. It doesn't really make a difference if 17 or 18 support requests were received.
At the same time, they would be somewhat forced to concede those use cases if they wanted to stick to the copyright story. Maybe you're right but I remain skeptical.
Yes but if the concern is being forced to remove code in the case of copyright violation it is trivial to remove code that is a test case as opposed to a feature of the compiler
Pangram is witchcraft to me. The way they can correctly detect AI writing from small samples, with so little statistical signal, is crazy. I've seen it correctly detect AI writing even when people use those "humanizer" rewriting skills that remove the hallmarks of AI writing and make the text indistinguishable from human text. But not to Pangram.
> I've seen it correctly detect AI writing even when people use those "humanizer" rewriting skills
Those skills are meant for humans, not for models. Pangram said in their recent model update that they now also detect the humanizers.
All it requires on their end is running each of their AI outputs through each humanizer and including it in their corpus as ai-humanized.
It's essentially impossible now to prompt your way to non AI detectable text. Even if you do today, it'll be picked up by the next model update.
This is why in academic environments its probably worth re-testing old exams or papers periodically - same way blood and urine samples from athletes are preserved to take advantage of better future testing.
Consider that they are in effect detecting whether text has been generated by one of a few dozen entities that has produced more text for them to analyze than any given human author.
Now imagine if they put the same effort into detecting if a given text was produced by one of a few dozen super-prolific human writers. I'd imagine they'd get pretty good at that too.
The main limitation of those "humanizer" rewriters is that most of them focus on making the next less detectable to humans, by making them read better. There's likely to be plenty of signal left that isn't affected by trying to make the text read better the same way human writers have plenty of idiosyncrasies despite being human.
I agree. I ran it on a small sample last night, basically three paragraphs of AI text that I thought read 100% human, except for em dashes, which I use in writing anyway. It flagged the first and last paras as "high" AI and the middle para as "high" human. I have no idea what it is seeing that sets it off.
This is the most important part most people don't get:
> And what would 3 °C mean for Germany?
> FB: In summer, meteorological records could reach up to 50 °C. Three degrees of global warming does not mean hot days will just be 3 or 4 degrees hotter. It could mean up to 10 degrees hotter. We would also face much longer droughts.
3°C warming implies summer days can get 10°C hotter. This is nightmare scenario.
It was always explained that way to me though. the "Our house is burning, and we are looking away" speech was like 20 years ago and it was pretty clear then than average increase in temps mostly mean higher maximums and minimums.
Nowadays, people who never did anything to consume less (and made fun of people who stopped taking the plane) can't hide behind the "we didn't know" because everybody knew, so they hide behind the "it was not predicted to be as bad" and other "the scientists explained that poorly", but we know what's up, even when they lie to themselves.
What do you mean? Both Europe and America moved industrial production to China (which was an explicit choice/policy) which as a side effect radically increased the amount of coal used to power industrial production, BUT lets Europeans and Americans both pay less (ie. make workers poorer) AND lets them claim innocence for CO2 increases, while in fact increasing their CO2 use indirectly.
Do you think that if China had isolated itself and refused all trade, it would have tried to avoid industry and emissions?
Do you think that China is controlled by its customers, making it a passive bystander in its own emissions, the agency and blame for which thus belongs to the customers?
Do you think that individuals who deliberately choose to emit less exist in significant numbers in China, or in the world, or ever might do?
That 3°C is about the yearly average of global temperatures, from deserts to poles, night and day, all year round, and maintained for several years. That is not weather, but it’s more like system energy.
Now, some regions get more average heating than others, so in that is not even, and then there is the weather, what you feel day to day. And right now we are already getting extreme heatwaves and other extreme weather events like flooding rains and hurricanes in zones where they were not usual. Those will increase in frequency and intensity, and I would not be surprised if there are peaks of temperature beyond 55°C, even before 2050, in industrialized countries
this also means thermal turbulence which means breakup of flow, and that means failure of distribution, leading to localized torrid spots, that translates to heat dome regions of more than 10°C hotter.
Are you being intentionally obtuse? In normal language the quoted text means there is a meaningful chance of record temperatures 10 degrees hotter than current records.
It is weasel language which actually says nothing in order to not be disprovable
The reason we see it all the time is because we've gotten used to people trying to convince us that some unlikely scenario will take place to increase urgency.
While global warming is happening, the quoted sentence should be raising your "this is someone trying to convince me of something that has a neglectable likelihood of becoming reality"-alarm
But sure, it could be that we may get a significant heat up within the next 20 years. Regardless of that, if you're living in Germany you probably should get an AC, as even if that doesn't manifest, things can already get too warm
When I clicked, I thought that it was going to smash the turbine blades from the animation to suggest something like "all the performance without the thermals," but nope, they just became laptop fans. And seconds later, it started blowing heat/steam out of them!
That's the most uncanny marketing for an ARM laptop I've ever seen.
When they exist at all. Many apps that provide important notifications (like delivery tracking, drop-off time etc) put them under the same category as marketing stuff. You can't have just the transactional tracking, you have to opt-in for the marketing notifications as well.
The ridesharing apps are the most annoying about this. Yes I want to be notified when my uber driver is almost here to pick me up. No, I don't want a notification about yet another sale.
It baffles me that they do this. I have to disable push notifications from Lyft entirely, so instead they send me ride updates as text messages, which surely must cost them way more money. Why not just introduce a "ride updates only" push notification category and stop this madness?
> It needs to be enforced by the OS or by law. Like how you get transactional emails without getting marketing spam.
What glorious universe do you live in where email is respected enough to have transactions separate from marketing, and that this is not only required by law but also enforced?
There's a pretty healthy regulatory environment around it, though I have noticed a resurgence of opt-out marketing communications on signup forms, which is unwelcome (I don't know if some legal decision changed, but it seemed like for a while this was not allowed, and maybe something has made companies think that it is again).
(I decided to look it up, here's the UK rules: https://ico.org.uk/for-organisations/direct-marketing-and-pr... . It looks like it is allowed to be opt-out if you buy something from them, which I do dislike, but there are rules and the ICO does have teeth)
Not sure about your experience, I’ve almost never encountered a marketing email that didn’t have an unsubscribe link, as is mandated by law in some countries. So I’m not really sure what you’re talking about.
On iOS atleast, Live Activities are separate from Notifications. So I can still monitor food or grocery delivery even though I have turned off their notifications.
Now a few apps have started sending notifications through WhatsApp because they have my phone number. e.g. Amazon
Yeah, but I still see apps that don't implement those features. Mostly React Native/Flutter apps that don't bother implementing native features. On Android it's even more depressing.
I'm not worried about missing food notifications because they send me an email and a text (... and a fax and a hardcopy confirmation letter in the mail.)
I had to disable from Android settings all LinkedIn notifications.
I check it from time to time but I haven't missed anything, nowadays LinkedIn is mostly garbage
Sure, but they didn't spend on training the model. If DeepSeek is providing the model for the same price as third parties, then it's probably still losing money when you account for the training.
Deepseek bypasses CUDA and has a few other optimisation that neither llama.cpp or vLLM support.
Furthermore, V4 pro was designed to run on 4 Huawei Ascend GPUs which are much cheaper than the nvidia setup others use, and deepseek probably also got some free hardware for their collab.
Hence it is entirely possible their inference costs are significantly lower than other providers.
Sometimes the person asking just needs an estimate, a ballpark number. It doesn't really make a difference if 17 or 18 support requests were received.