There's a significant quality difference however if you look at the generated samples in the paper. Imagen Video is leagues ahead. The progress is still quite drastic
There's actually a dating app with a similar-ish concept called Happn (https://happn.com/), which basically is Tinder but if Tinder only showed you the profiles of people you had actually crossed paths with in real life. Fun fact! Not exactly what you said but close!
Online courses? Do you mean "investing in yourself" per se? Any ideas for growing the money that involve actual investment instruments, if I hypothetically have $10k lying around?
The ballotpedia summary shows that there wasn't really a campaign opposing the measure it seems? In which case presumably it makes sense that it passed overwhelmingly, if no one was voicing any arguments against it
What is the biggest Ada-based project currently in use? Does anyone know? Is Ada more for DoD-type backend-stuff, where Fortran or C would be used, or something else? What is Ada for?
For anyone looking for schools for CS, https://csrankings.org/ is an excellent site that gives you an overview of departments' research output, filterable by subfield, geographic location etc. A very nice system if you're intending to get involved in research in your undergrad, but especially good if you plan on continuing to a Master's or PhD in CS.
On the flipside, natural language interfaces have the potential to be extremely easy to use for anyone, including non-experts. Anyone can type a message to the computer, without having to learn the specifics of an interface's custom controls. There are different types of efficiency. I'm assuming you're referring to something like 'operational efficiency', while NLI wins on 'ease of adoption' per se.
How exactly do game engines monetize, can anyone shed any light? Is it via additional support in using the engine for large companies? A freemium model like Unity's? But didn't even Unity succumb eventually to supplementing monetization via ads?
It's interesting, because through my readings in linguistics, I got the sense that no language is "objectively" harder than any other, that everything is relative to the other languages you know. A Dutch speaker will learn German potentially more easily than French compared to an English speaker, so how do you judge objectively which language is harder? Nevertheless, an entertaining read! I have heard a good candidate for the hardest natural language is Georgian, due to irregular grammar, long consonant clusters, and difficult script. Not sure how to reconcile the apparent objectively difficult features with the previous claim. I guess in the end difficulty might just be a direct result of number of (relatively) unique features compared to other languages (of which they say Georgian has many)
Is there any point on which Tensorflow actually wins out against Pytorch? My sense is that Pytorch is easier for almost every if not every single use case. Not to mention the 1 => 2 transition gave me serious Python 2 => 3 vibes with the switch to default eager execution etc. Anyone still using Tensorflow care to weigh in?
How long does it take to generate a single image? Is it in the 30 min type range or a few mins? It's hypothetically "possible" to run e.g. OPT175B on a consumer GPU via Huggingface Accelerate, but in practice it takes like 30 mins to generate a single token.
Oh shoot, I thought it was for everything. Fair enough. Although I think new competitive features like Outpainting are definitely in response to Stable Diff
Damn, Dall-E really lost its competitive edge overnight when Stable Diffusion was released. They dropped their prices across the board in response, but honestly I think it still isn't enough to save them. The magic of open-source competition.
Just as photoshop became another tool to use in a designer's library, so to will it be with image generation (a la Stable Diffusion, DALL-E, etc.). There's no reason to think it'll be any different. Sure, the skillset will shift, learning to design prompts will become important for any designer worth their salt, but the job of designer isn't going away. It'll just be changed, just as photoshop changed it and provided new skills to learn and leverage.
In the end, isn't the answer relatively simple when phrased in terms of supply and demand? In by far the majority of these cities, the number of new units built pales in comparison to the number of new people moving to these cities, and as such you have a larger number of people competing for the same number of units, thus driving up prices. Cities like Montreal, Canada have a much larger housing stock due to middle density housing (multiplexes of 6-10 apts), and this is reflected in these cities having much lower average rent compared to other cities in North America.
Honestly Lua is one of those languages that seems decently nice to use, fairly performant, and just makes me wonder why it hasn't been adopted for use more broadly in domains beyond just like being embedded in game engines. Am I missing something?
Honestly? I try not to think about the future because doing so brings wave upon wave of existential dread. Things aren't looking good, and it might be a bit of a hedonistic maxim but if thinking about it just makes me feel worse why should I? Let me work away at achieving my "local optimum" since that's all I can realistically do.
You're right in that the term Hellenes appears later, however it is worth noting that there was still some sense of collective identity, even that far back, as evidenced by the tribes coming together when facing external threats (e.g. the Persians, or even against Troy in the Iliad), common language, culture, competitions that everyone participated in, etc. etc.
Something I always wondered: the Gamecube had an IBM PowerPC processor with a RISC instruction set. ARM is also a RISC instruction set. Is there some performance boost in running Dolphin on ARM-powered hardware, or is it more or less the same as running it on an x86-based chip? My sense is that RISC -> RISC potentially would be more straightforward?
actually sorry to be pedantic but it's legomenon not logomenon, it's a participle of the verb λέγω, "to say", not λόγος (reason), which I'm assuming explains the mistake.
the framerate dropped in certain areas, it ran mostly consistently at 30fps but not perfectly consistently. These mods make it so that even in those areas it still maintains 30fps.
This makes sense, OpenAI has been experimenting with human-in-the-loop type refinement systems for GPT-3 (see InstructGPT https://openai.com/blog/instruction-following/). It would make sense that they would use something similar for Dall-E 2 as well.