Unless someone convinced someone that they should prepared in case they become BIG and get millions of concurrent users.
And even then, nothing is sure. During the COVID-19 pandemic, Pokémon Go, thanks to an adaptation of game mechanics, literally exploded in users. Despite being world-scale, they struggled for a couple of weeks to meet the workload. World-scale naturally degrades to normal-scale if you don't stress it regularly.
> That would not be appropriate in any of the mainstream cloud-native architecture styles.
I mean it ends up being like that. A significant number of medium sized sites bust out into micro-services at some point. Or they start off as serverless, and realise that actually its not as easy to scale as it was claimed.
It isn't just the projections that distort our perception. North being up and south being down is so ubiquitous that it seems like Earth (and the Solar System) has a top side and a bottom side. But that's just a convention.
Unfortunately it follows that with "occasionally connect the TV to the internet for a minute to see if it needs any firmware updates" which is pointless if the TV is already working properly.
Possibly. Judging only by my own (heavily Asian) neighborhood, little white dogs gave way to (Goldens / Labs and -doodles). I can't even remember all the yellow Labs' names anymore.
The training data is so large that it incidentally includes basically anything that Google would index plus the contents of as many thousands of copyrighted works that they could get their hands on. So that would definitely include some test prep books.
They seem to be taking this into account: We did no specific training for these exams. A minority of the problems in the exams were seen by the model during training; for each exam we run a variant with these questions removed and report the lower score of the two. We believe the results to be representative. (this is from the technical report itself: https://cdn.openai.com/papers/gpt-4.pdf, not the article).
By the same token, though, whatever test questions and answers it might have seen represent a tiny bit of the overall training data. It would be very surprising if it selectively "remembered" exact answers to all those questions, unless it was specifically trained repeatedly on them.
If the future of AI is LLMs like ChatGPT, which are trained on literature and other things that people create, you're going to need humanities scholars like you need computer scientists to understand the AI. Microsoft gave their chatbot, which has probably almost every published work of science fiction in its training set, a human name and then were surprised that it imitated the fictional poorly-behaved named AIs that it was exposed to in its training.
I've been spending some time trying to get a sense of how it works by exploring where it fails. When it makes a mistake, you can ask questions in a socratic method until it says the true counterpart to its mistake. It doesn't comment on noticing a discrepancy even if you try to get it to reconcile its previous answer with the corrected version that you guided it to. If you ask specifically about the discrepancy it will usually deny the discrepancy entirely or double-down on the mistake. In the cases where it eventually states the truth through this process, asking the original question that you started with will cause it to state the false version again despite obviously contradicting what it said in the immediately previous answer.
ChatGPT is immune to the socratic method. It's like it has a model of the world that was developed by processing its training data but it is unable to improve its conceptual model over the course of a conversation.
These are not the kinds of logical failures that a human would make. It may be the most naturalistic computing system we've ever seen but when pushed to its limits it does not "think" like a human at all.
> If you ask specifically about the discrepancy it will usually deny the discrepancy entirely or double-down on the mistake.
I have had the exact opposite experience. I pasted error messages from code it generated, I corrected its Latin grammar, and I pointed out contradictions in its factual statements in a variety of ways. Every time, it responded with a correction and (the same) apology.
This makes me wonder if we got different paths in an AB test.
I pasted error messages from code it generated. It kept generating the same compiler error eventually.
When I applied the "socratic method" and explained to it the answer based on stack overflow answers. It would at first pretend to understand by transforming the relevant documentation I inserted into it, but once I asked it the original question, it basically ignored all the progress and kept creating the same code with the same compiler errors.
That would not be appropriate in any of the mainstream cloud-native architecture styles.