> This is a typical tech echo chamber. There is a significant number of people who make direct purchases through ads.
Even if its not, TV ads, newspaper ads, magazine ads, billboards, etc... get exactly 0 clickthrus, and yet, people still bought (and continue to buy) them. Why do we act like impressions are hunky-dory for every other medium, but worthless for web ads?
I don't see why, unless you think they're lying and they filmed their demos, or used some other preexisting model. I didn't ignore the JWST launch just because I haven't been granted to ability to use the telescope.
Back when Imagen was not public, they didn't properly validate whether you were a "trusted tester" on the backend, so I managed to generate a few images..
..and that's when I realized how much cherry picking we have in these "demos". These demos are about deceiving you into thinking the model is much better than it actually is.
This promotes not making the models available, because people then compare their extrapolation of demo images with the actual outputs. This can trick people into thinking Google is winning the game.
Her approach might not be the literal peak healthiest, but that doesn't mean its wrong. If she's only having a drink every once in a while and is getting 7+ hours of sleep most nights then the negative effects are probably negligible anyway.
To a certain extent, most people need to sacrifice some amount of health and performance to live the fullest life they can. If you don't feel that you benefit from that, though, and its important for either of you that you share your lifestyle, then perhaps you just weren't that compatible.
> To a certain extent, most people need to sacrifice some amount of health and performance to live the fullest life they can.
I agree 100%. I did not post to complain about her, but more to note about incompatibilities in things like sleep schedule. I do believe though she should drink a bit less, but as long as it makes her enjoy life, why not.
Not the person you're responding to, but no, it doesn't really bother me at all. What does bother me is that I don't have confidence in the value of the output, where as if I listen to This American Life, or a podcast or audiobook from a trusted authority, I don't have to worry about that.
I don't doubt you're right about social media and smartphones rotting our attention spans. But also, peripatetic philosophy is ancient. I spend most of my day sitting. Whether its work, entertainment, or hobbies, most of these things have me sat in front of a screen. So its nice, and I do think it increases my retention, to be able to do something while walking or cycling instead of sitting.
While you're right that the Internet contains a lot of wonder and exploration, the vast majority of people (and kids) will not interact much with that part of the Internet, if at all. Additionally, social media platforms have collapsed what would have been standalone, somewhat magical experiences into their own uniform platforms. I've heard someone say that kids today tend to think in terms of "apps" and not "websites", because rather than having everything scattered across a lot of small, independently maintained, websites, there are instead a few web apps that contain 99% of what you want to get at.
That means that if you really want to "surf" the web these days you have to dig deep and avoid getting sucked into a social media platform. And when you do dig deep there's not that much out there, because the people who would be maintaining their own web page now just have a facebook page for their business and a twitter account for their personal posts.
> Many of these alternatives failed because they were forced onto users who had already paid over $60 for a game
I really doubt that hurt their adoption. Yes, it pissed people off, but that doesn't mean it suppressed adoption. Being slow and clunky, sure, but you're probably not going to get anywhere without some high value exclusives.
Extremely pedantic, but is "non-deterministic" really the right language? The same input will always produce the same output, provided you haven't intentionally configured the system to use the model non-deterministically. It seems like the right way to describe it is as a chaotic deterministic system. The same input will always produce the same output, but small shifts in the input or weights can result in dramatic and difficult to predict changes in outputs.
> The same input will always produce the same output
Not guaranteed even with the same seed. If you don't perform all operations in exactly the same order, even a simple float32 sum, if batched differently, will result in different final value. This depends on the load factor and how resources are allocated.
Yeah, the fact that floating point multiplication isn't associative is a real pain for producing deterministic outputs - especially when you're running massively parallel computations on GPUs (or multiple GPUs) making the order of operations even less predictable.
Even if its not, TV ads, newspaper ads, magazine ads, billboards, etc... get exactly 0 clickthrus, and yet, people still bought (and continue to buy) them. Why do we act like impressions are hunky-dory for every other medium, but worthless for web ads?