> If AI use produces obviously inferior work, how did it win in the first place?
they uses some AI placeholders during development as it can majorly speed up/unblock the dev loop while not really having any ethical issues (as you still hire artists to produce all the final assets) and in some corner case they forgot to replace the place holder
also some of the tooling they might have used might technically count as gen AI, e.g. way before LLM became big I had dabbled a bit in gen AI and there where some decent line work smoothing algorithms and similar with non of the ethical questions. Tools which help removing some dump annoying overhead for artists but don't replace "creative work". But which anyway are technical gen AI...
I think this mainly shows that a blank ban on "gen AI" instead of one of idk. "gen AI used in ways which replaces Artists" is kinda tone deaf/unproductive.
> AI placeholders during development as it can majorly speed up/unblock
Zero-effort placeholders have existed for decades without GenAI, and were better at the job. The ideal placeholder gives an idea of what needs to go there, while also being obvious that it needs to be replaced. This [1] is an example of an ideal placeholder, and it was made without GenAI. It's bad, and that's good!
A GenAI placeholder fails at both halves of what a placeholder needs to do. There's no benefit for a placeholder to be good enough to fly under the radar unless you want it to be able to sneak through.
it's not better as they fundamentally fail to capture the atmosphere and look of a scene
this means that for some use cases (early QA, design 3D design tweaks before the final graphic is available etc.) they are fully useless
it's both viable and strongly preferable to track placeholders in some consistent way unrelated to their looks (e.g. have a bool property associated with each placeholder). Or else you might overlook some rarely seen corner cases textures when doing the final cleanup
so no, placeholder don't need to be obvious at all, and like mentioned them looking out of place can be an issues for some usages. Having something resembling the final design is better _iff_ it's cheap to do.
so no they aren't failing, they are succeeding, if you have proper tooling and don't rely on a crutch like "I will surely notice them because they look bad"
I've actually considered hiring artists to help me out a few times too under sort of comparable circumstances? I could use AI to generate basic assets, and then hire artists for the real work! More work for artists, better quality for me. Unfortunately, I fear I'd get yelled at (possibly as a traitor to both sides?)
Frankly, in the wider debate, I think engagement algorithms are partially to blame. Nuanced approaches don't get engagement, so on every topic everyone is split into two or more tribes yelling at each other. Folks in the middle who just want to get along have a hard time.
(Present company excepted of course. Dang is doing a fine job!)
it's an (maybe the most) extreme example of something which is "gen AI" but not problematic
and as such a naive "rule" saying "no gen AI at all" is a pretty bad competition rule design
Reductio ad absurdum is a form of logic which takes an argument to its logical conclusion in order to demonstrate that it is absurd if it were to be taken on its face. Whether or not anyone applies it that way in reality is irrelevant.
some pretty decent "line smoothing" algorithm are technically gen AI but have non of the ethical issues
You'd have to cite an actual example of something this ridiculous. Non gen-AI algorithms have been line smoothing just fine for 2 decades now for less than a trillionth of the resources required to use gen AI for the same task.
yes, and here is a fun fact, most of the push for mass surveillance comes from the European Council, the thing is that literally are "just" the locally elected leaders...
not some vague far away "the EU (personalized)" thing
which also mean you can locally enact pressure on them
furthermore the EU supreme court(s) might have more often hindered mass surveillance laws in member states then the council pushing for them...
and if we speak as of "now", not just the UK, but also the US and probably many other states have far more mass surveillance then the EU has "in general".
so year the whole "EU is at fault of everything" sentiment makes little sense. I guess in some cases it's an excuse for people having given up on politics. But given how often EU decisions are severely presented out of context I guess some degree of anti-EU propaganda is in there, too.
> mass surveillance comes from the European Council, the thing is that literally are "just" the locally elected leaders...
Factually incorrect.
The European Parliament is elected. The Council is appointed, so there is no direct democratic incentive for the council to act on and no direct electorate to please.
On top of that the actually elected European Parliament can only approve (or turn down) directives authored by the Council. They have no authority to draft policies on their own.
To make matters even worse the European Council, which drafts the policies, has no public minutes to inspect. Which obviously makes it ripe for corruption. Which evidently there is a lot of!
Looking at the complete picture, the EU looks like a construct designed intentionally to superficially appear democratic while in reality being the opposite. The more you look at how it actually works, the worse it looks. Sadly.
In short, there are three core institutions, the "technocratic" European Commission, the European Parliament elected by direct popular vote, and the Council ("of the EU"/"of ministers") made up of the relevant (in terms of subject matter) ministers of the standing national govs. The law-making procedures depend on policy areas etc. but usually in the policy areas where EU is fully competent, the Commission — the democratically least accountable of the three bodies — by default makes the initiatives and negotiates/mediates them further along with the Parliament and Council, but only the last two together really have the power to finally approve actual legislation, usually either Regulations (directly applicable in member states as such — so an increasingly preferred instrument of near-full harmonisation), or Directives (requiring separate national transposition / implementation and usually leaving more room for national-level discretion otherwise as well).
While not fully comparable to nation-state parliaments, the powers of the EU Parliament have been strengthened vis-à-vis both the Commission and the Council, and it's certainly long been a misrepresentation to say that they, e.g., only have the power to "approve or turn down" proposals of the Commission and/or the Council.
it was the EU which had stopped many similar unhinged attempts from the UK when the UK was still a member
similar it had been the EU which had shut down various other surveillance nonsense of the EU
you are basically pretending the EU is a person with one uniform opinion and goals
but it's like the opposite of it, like in a lot of way
it's a union of states, each having a vastly different goals and culture and non of them having a "single uniform opinion" either but (in most cases) a more complex political field then the US (on a federal level)
Furthermore the most influential organ of the EU when it comes to making changes is literally a composition of the elected leaders of the member states. So for most big controversial decisions the driving and directing force isn't "the EU" but but the various elected leaders of the member states. For EU citizens blaming "the EU" instead of blaming your own elected leaders is common, but pretty counter productive, as it's basically pretending you have no power to change things.
Furthermore in the EU you have an additional parliament which (in general) needs to ratify laws and two high courts which can (and in context of mass surveillance repeatedly have) shut down misguided "laws", including in many cases local attempts at mass surveillance laws.
So while some parts of the EU have consistently pushed for mass surveillance in recent years other parts also have consistently moved against it.
In general while the EU needs a lot more transparency and some more democratic processes in some aspects a lot (not all) of the "stories told to make the EU look dump/bad" have a lot of important context stripped from that (like e.g. that a lot of the current push for surveillance comes from the locally elected leaders not the EU parliament or some other abstract "the EU" thing, it's your own countries leader/lead party(1) which does or at least tolerates that shit).
> blaming "the EU" instead of blaming your own elected leaders
The elected leaders like to blame the EU (or for those without an EU - any external body or even the mythical deep state) for everything adverse. The reality is these "failures" they blame on someone else are generally in alignment with their own policies goals and objectives.
for quite a while I through many of those dump "internal network scanning automatized pentests" where pretty pointless
but after having seen IRL people accidentally overlooking very basic things I now (since a few years) think using them is essential, even through they often suck(1).
(1): Like due to false positives, wrong severity classifications, wrong reasoning for why something is a problem and in generally not doing anything application specific, etc.
I mean who would be so dump to accidentally expose some RCE prone internal testing helper only used for local integration tests on their local network (turns out anyone who uses docker/docker-compose with a port mapping which doesn't explicitly define the interface, i.e. anyone following 99% of docker tutorials...). Or there is no way you forget to set content security policies I mean it's a ticket on the initial project setup or already done in the project template (but then a careless git conflict resolution removed them). etc.
This article https://signal.org/blog/building-faster-oram/ has some details but is more focused on improving their solution other blogs from the are "we want to build this soon" kind of blogs. It seems that most articles about this topic either have too little content to be of interest or are technology previews/"we maybe will do that" articles about things Signal wants to implement, where it's unclear if they did do that or something similar.
To cut it short they use Intel SGX to create a "trusted environment" (trusted by the app/user) in which the run the contact discovery.
In that trusted environment you then run algorithms similar to other messengers (i.e. you still need to rate limit them as it's possible to iterate _all_ phone numbers which exist).
If working as intended, this is better then what alternatives provide as it doesn't just protect phone numbers from 3rd parties but also from the data center operator and to some degree even signal itself.
But it's not perfect. You can use side channel attacks against Intel SGX and Signal most likely can sneak in ways for them to access things by changing the code, sure people might find this but it's still viable.
In the end what matters is driving up the cost of attacks to a point where they aren't worth in all cases (as in either not worth in general or in there being easier attack vectors e.g. against your phone which also gives them what they want, either way it should be suited for systematic mass surveillance of everyone or even just sub groups like politicians, journalists and similar).
Can someone please clarify: For the phone number to reach the enclave for use during search via XOR, won't it need to come in as regular RAM via the backend's API call?
the problem only affect a subset of HDMI 2.1 features, not HDMI 2.0
but the steam machine isn't really super powerful (fast enough for a lot of games, faster then what a lot of steam customers have, sure. But still no that fast.)
So most of the HDMI 2.1 features it can't use aren't that relevant. Like sure you don't get >60fps@4K but you already need a good amount of FSR to get to 60fps@4k.
Just because the Steam Machine isn't powerful enough to support high framerates in modern AAA games doesn't mean it can't do so with older or less graphically-intensive games.
VRR and HDR are presumably the biggest issues, because HDMI 2.0 should already have enough bandwith to support 8-bit 2160p120 with 4:2:0 chroma subsampling, which should work fine for most SDR games, and 144 Hz vs 120 Hz is, in my experience at least, not noticeably different enough to be worth fussing over.
Some people will want to use their Steam Machine as a general-purpose desktop, of course, where RGB or 4:2:2 is nonnegotiable. Though in this case 120 Hz — or 120,000/1001 Hz, thanks NTSC — is, again in my experience, superior to 144 Hz as it avoids frame pacing issues with 30/60 Hz video.
A better approach would be to put your energy into making sure the
used methods are _reasonable_.
We don't require every FSK16 game sail to register the buyers name,
age, contact info on physical checkouts etc. In most countries a law
requiring that would be seen as excessive and in some places unconstitutional.
Instead it's fine to visually look at a id, and if it "obvious" they are adult (e.g. very old person) we don't even require that. And thats fine. Because we don't need
a perfect prevention we just need something which helps parents parent "a bit" and helps "a bit" in cases where parents don't parent.
If everyone fight "all age check solutions" the chance that they get fully ignored and some horrible shit gets passed into law is very high.
If everyone fighting also provides a alternative and strict guidelines about what is and isn't acceptable in their opinion there is a chance for reasonable solutions being implemented instead.
(Like e.g. put a age gate header into http responses, like "min-audience-age: region=US, age=123; region=EU, age=456", say OS must have a API where you pass that in and they say yes/no for that account, do NOT require any crypto, signing etc. This is not fraud prevention but parenting helper. The OS then can store `18+|age` internally and have some integrations with country specific age verification services (it must only store 18+|birthday and only birthday iff <18, I guess for US 21). But there is no need to prevent anyone from changing this value with e.g. windows regestry changes, except if it's a child account. So require any widely _sold_ OS to have a parent controls/child account functionality.
But really any solution which effectively requires mass surveillance, exclude hobby OS or similar, require some clever signing scheme involving device attestation etc. is VERY excessive and unneeded.
What I don't understand (but also wouldn't be surprised about if it is misrepresented by the article) is:
- why would you get a single, for ADHD, non-social-related anxiety, non-sever autism or depression (especially in the later case you probably shouldn't be in a single)
- I mean sure social anxiety, sever autism can be good reasons for a single.
through in general the whole US dorms thing is strange to me (in the EU there are dorms, but optional (in general). And 50%+ of studentsfind housing outside of it (but depends on location). This allows for a lot more individualized living choices.)
they uses some AI placeholders during development as it can majorly speed up/unblock the dev loop while not really having any ethical issues (as you still hire artists to produce all the final assets) and in some corner case they forgot to replace the place holder
also some of the tooling they might have used might technically count as gen AI, e.g. way before LLM became big I had dabbled a bit in gen AI and there where some decent line work smoothing algorithms and similar with non of the ethical questions. Tools which help removing some dump annoying overhead for artists but don't replace "creative work". But which anyway are technical gen AI...
I think this mainly shows that a blank ban on "gen AI" instead of one of idk. "gen AI used in ways which replaces Artists" is kinda tone deaf/unproductive.
reply