Joined: 11/13/2006
Posts: 2924
Location: Northern California
I brought this up about a week ago and I think it's finally time to bring it to the public for discussion.
A little over a month ago, prompted by internal discussion over this issue on BizHawk's Github page, I drafted up a set of AI usage policies for TASVideos. It was in a slow process of being discussed and refined internally amongst senior staff, but a recent submission openly relying on ChatGPT assistance hit the site and the negative response to it from the community made me realize that we weren't really going fast enough with it.
So, I'd like to present the most recent version of my draft to the community for public discussion: Wiki: HomePages/Samsara/Scratchpad/Draft2UPDATE (August 1st, 2026): Brand new, fully rewritten draft just dropped. For comparison purposes, the original is still located here.
I've run this by the rest of staff already, who have pitched in their thoughts and suggestions on what could be changed and improved, but I would still like to stress that this policy is primarily my own thoughts and feelings and is not meant to be taken as the site's official policy as of yet. For something like this, I want to make sure as many people as possible are on board with it, that ideally everyone agrees with it and understands my reasoning. I did my best to account for every possible possibility and nuance ahead of time, though with this opening up to public discussion I'm sure there are still going to be plenty of questions I haven't considered yet.
The main thing I want feedback on is whether or not the balance is right, in that I want to make sure that any lenience I have chosen to give in defiance to my own vehemently anti-LLM beliefs is reasonable for the community at large. If you all aren't comfortable with it, then I can't in good conscience let it go into effect until you are. In that sense I will almost always trend towards shifting the balance towards further restriction, and will most likely never shift it towards further lenience.
Discussion will take place both here and on Discord until I (and ideally the rest of staff) feel as though the community is fully on board with the policy.
Just as a comment on terminology: The policy proposal mostly uses the term LLM, but AI models that generate non-text materials like images and videos aren't LLMs. "Generative AI" could be a better term to use to make it clearer that those are also covered.
Joined: 11/13/2006
Posts: 2924
Location: Northern California
JosJuice wrote:
Just as a comment on terminology: The policy proposal mostly uses the term LLM, but AI models that generate non-text materials like images and videos aren't LLMs. "Generative AI" could be a better term to use to make it clearer that those are also covered.
That's briefly addressed here. Outside of calling it an AI policy, I tried to avoid using "AI" as a term just to differentiate our standard usage of it from the LLMs/GenAI that the policy is about. I can maybe swap the media generation usages from LLM to GenAI, though I'd prefer a different term that doesn't use "AI" in it at all just to keep things as clear as possible. Ideally, "AI" wouldn't have been co-opted and misused to this degree, and gaming communities like us wouldn't need to worry about possible confusion as a result, but unfortunately this is far from an ideal world.
You're right, I AM overthinking this!
i don't understand what is going on, why would we be opposed to using AI or LLM or whatever it is to make TASing easier? why wouldn't Igor's work be acceptable for using AI? If ChatGPT could help me TAS something, i would 100% use it to make it faster or even reach results i can't with my current skillset or knowledge.
I want all good TAS inside TASvideos, it's my motto.
TAS i'm interested:
Megaman series, specially the RPGs! Where is the mmbn1 all chips TAS we deserve? Where is the Command Mission TAS?
i'm slowly moving away from TASing fighting games for speed, maybe it's time to start finding some entertainment value in TASing.
Plagiarism is unacceptable even if it's through a machine.
[16:08:10] <BenLubar> a TAS is just the limit of a segmented speedrun as the segment length approaches zero
[3:56 PM]Thursday, July 23, 2026 3:56 PM Samsara: i mean this in the most positive way possible
[3:56 PM]Thursday, July 23, 2026 3:56 PM Samsara: i feel like a lot of your interaction with tasvideos is finding excuses to watch TASes
i don't understand what is going on, why would we be opposed to using AI or LLM or whatever it is to make TASing easier? why wouldn't Igor's work be acceptable for using AI? If ChatGPT could help me TAS something, i would 100% use it to make it faster or even reach results i can't with my current skillset or knowledge.
Did you read the proposed policy, Igor's submission thread, or any of the other links from Samsara's post before asking these questions?
Have you considered the very real costs and consequences of using this technology before declaring you'd absolutely use it?
Joined: 11/13/2006
Posts: 2924
Location: Northern California
KusogeMan wrote:
i don't understand what is going on, why would we be opposed to using AI or LLM or whatever it is to make TASing easier? why wouldn't Igor's work be acceptable for using AI? If ChatGPT could help me TAS something, i would 100% use it to make it faster or even reach results i can't with my current skillset or knowledge.
We recognize that LLMs are a tool, and like any other tool it is possible to use them for beneficial purposes, particularly in accessibility. Some users, such as those who are disabled or those who do not speak or understand English, may rely on LLM text generation in order to be able to communicate on the site at all. As much as we wish these tools could be used in much better circumstances, we have chosen to allow them as a way to ensure accessibility to the site is as open as possible.
While it is almost certainly impossible to generate a working input file with an LLM, let alone an optimized one, it is still possible to use LLMs for assistance in TASing through helping with analysis and code writing. It is, in theory, also possible for AI-generated assets to be used in an arbitrary code execution run. Due to these possibilities, we are treating these usages the same as any other contextual usage: We strongly discourage, but will still hesitantly allow, text generation in research and analysis,
As morally opposed as I am to the industry, and as confused as I am that people are actively arguing their right to take all the fun and creativity out of the creative and fun hobby that they do to have fun and express their creativity, I still went out of my way to say that using it as a tool is still acceptable here, and the rest of staff agreed with me.
I want to make this clear, again, even though I thought I already did several times in the policy: This isn't about the tech itself. All the arguments I've seen against the policy have focused entirely on the tech itself, which is misrepresenting our stance to a level that I can only view in bad faith, whether or not it's intentional. I don't think anyone here is against the actual tech. To be honest, I'm personally fascinated by it. I even set up local environments for it and played around with it as a personal interest a couple years back. I think it's genuinely incredible. The problem is that the current state of the industry makes any usage of the tech inherently and unavoidably unethical and damaging in its current state. If genAI was regulated, if it wasn't literally killing the planet and literally killing people, if it was ethically sourced from consenting individuals that believe in the product and controlled in a way that didn't allow bad actors to profit off of planetary-scale plagiarism, I doubt anyone in the world would have major concerns over it. If the industry wasn't a massive, money-obsessed shitshow that has to mislead and manipulate people in order to stay afloat, this policy wouldn't exist, nor would any other.
To be more clear and more blunt about it, I'm not interested in continuing to listen to people minimizing or outright ignoring our true concerns in order to fight for rights they already have. We have already allowed agentic assistance in TAS creation, as Walgrey pointed out above, and the policy is clear about continuing to allow it. I'm not even asking for disclosure with any text generation or agentic assistance, because I don't think it's warranted for TASing and would only lead to unnecessary infighting.
As a final note, keep in mind this policy wasn't prompted by the SMW submission. The initial draft was made before that submission, with plenty of easily accessible evidence to back that up. I just chose to expedite it to public discussion because of the backlash. There are some elements that were written as a result of the submission, of course, but those elements are consistent with the initial draft and vision. Most of what that submission did is allowed by the draft and has already been allowed in other submissions, the sole exception in my mind being the ACE'd in thank you to ChatGPT as a part of the actual run. I feel that crosses the line into outright endorsement that I am not comfortable showcasing as a representative of this community. There's a difference between just using agentic assistance and singing its virtues loud enough for the world to hear even over the sound of the water it's rapidly evaporating.
I read through the proposed policy, and I mostly agree overall with the current state of how it's written. However, I do feel like there may be additional restrictions warranted on the text generation side of things.
While I agree with the portions of the policy regarding LLM assistance with translation or improving personal writing style, I think this sort of usage should be restricted to instances where the user is speaking only for themselves (or their co-authors, in the case of submission notes). I think shared resources/pages that many users can edit should be off-limits for such usage. I've seen instances of people giving game-specific "advice" generated from LLMs that's just flat-out wrong, and this would certainly be undesirable on shared pages where the false LLM information could be mixed into the rest of the correct, human-written information.
Also, I think it's worth adding a note for the allowed usage of LLMs that users who use LLM assistance for writing posts are still personally responsible for ensuring that the generated text follows site rules. This is of course redundant, but emphasizing it here highlights the importance of actually checking the output before posting it and may prevent people trying to wring their hands of any wrongdoing by blaming the LLM for something inappropriate/plagiarized.
One thing I wanted to ask about: is the "media containing LLM-generated content" meant to encompass submissions of AI-generated games or games that use AI-generated assets? It might be worth explicitly clarifying that in particular, especially as the number of these games continues to increase.
Joined: 11/13/2006
Posts: 2924
Location: Northern California
OceanBagel wrote:
While I agree with the portions of the policy regarding LLM assistance with translation or improving personal writing style, I think this sort of usage should be restricted to instances where the user is speaking only for themselves (or their co-authors, in the case of submission notes). I think shared resources/pages that many users can edit should be off-limits for such usage. I've seen instances of people giving game-specific "advice" generated from LLMs that's just flat-out wrong, and this would certainly be undesirable on shared pages where the false LLM information could be mixed into the rest of the correct, human-written information.
This is a fantastic point and definitely one that should be accounted for given how often I've seen our resource pages linked in other communities. Should be very easy to add to the draft as well since it's the same principle as this. On the same line, it's also worth discussing whether or not the volunteer positions like Editor and Reviewer are considered staff-adjacent enough to need to hold themselves to the site policy. My gut says yes: Despite TASVideos being a wiki, most of the pages are staff-written and maintained, so any LLM usage on wiki pages would most likely be attributed to us even if it was added by non-staff. Given we've recently set up an application form for these positions as well...
Also, I think it's worth adding a note for the allowed usage of LLMs that users who use LLM assistance for writing posts are still personally responsible for ensuring that the generated text follows site rules. This is of course redundant, but emphasizing it here highlights the importance of actually checking the output before posting it and may prevent people trying to wring their hands of any wrongdoing by blaming the LLM for something inappropriate/plagiarized.
Also a great point, though a bit harder to put into practice since not everyone may be able to adequately check the output, particularly those using it for translation with zero fluency in English or whatever other language they're translating to.
One thing I wanted to ask about: is the "media containing LLM-generated content" meant to encompass submissions of AI-generated games or games that use AI-generated assets? It might be worth explicitly clarifying that in particular, especially as the number of these games continues to increase.
This is actually a harder question to answer than I thought it would be. The obvious answer is yes, it would apply, but at the same time more and more AAA companies and indie devs have been embracing (or are just forced to use) the tech to stay competitive and it might eventually reach the point where we just have to accept it as another unfortunate reality of genAI proliferation. I don't exactly know how much this affects us since a huge portion of our current and potential libraries are all pre-genAI, but someone did join the Discord a couple days back to shill their vibecoded NES homebrew with accompanying genAI instruction manual so the effect is technically already non-zero, and I'm sure it's only going to get worse as time goes on. GenAI will proliferate further, our own tech will get more accessible and reach newer games...
Gut reactions again here: I think there's a difference between "TASing someone else's game that uses agentic code" and "posting your vibecoded NES homebrew with accompanying genAI instruction manual". The latter sends you straight to jail, do not pass Go and ESPECIALLY do not collect $200 or any money at all you absolute slop sycophant, but the former definitely requires further nuance, and I can't really formulate anything clear and understandable right now. I definitely need to think about this further, and when I'm much less tired.
At the very least I should have the first two points accounted for in the draft sometime tomorrow. The last one absolutely needs some further discussion, but I'm sure we should all be able to agree on something concrete once that discussion happens. Thank you so much for bringing all of this up!
I have read through the policy, and it's just about perfect. I've been waiting for this for ages. It's just that there should be more of an emphasis on how unethical Generative AI is. Or to be more nuanced and specific, some specific (and a majority) of Generative AIs. A suggestion that I have is to increase(or decrease, I kinda forgot the grammar) the threshold for people using generative AI.
And also, it should also be known that Tool-Assisted Speedrunning is a form of art, so Generative AI usage in TASes should be more restricted than it currently is. Just like how AI generated art is very frowned on.
Some rambly thoughts on the topic that haven't coalesced enough in my mind to form a coherent take, but that I still felt strongly enough about to post here:
I am personally very conflicted on AI usage. Personally, I absolutely, unequivocally, loathe AI and the AI industry for all the reasons mentioned in the proposal (and a couple others not mentioned as well), but I also have to begrudgingly admit it can be a helper and an enabler for some people. Especially in a TASing context, there is still plenty of room for people to synthesize enough original work on top of an LLM's output to make a meaningful product, like
Using an LLM to dig into the game's internals to understand tricks or ACE exploits*
Vibe-coding a Lua script that aids someone in their TASing process.
I personally would not want to touch AI for these things, but to each their own.
There's also the possibility of using your own locally running models, or finding a model with ethically-sourced training data, which defuses some (but not all) of the problems with AI**.
I think as an AI-averse viewer, my primary two concerns are:
I'd appreciate transparency. If you are a TASer, I appreciate you explaining your TASing process and your tools. This has always been true, but I would like to be able to make a conscious choice if I want to support a particular instance of Gen. AI usage. Again, in the case of TASing, if you just used AI to write a Lua script for yourself, I think that's your prerogative. After all, if it doesn't materially affect the final input file, and unlike vibe-coded software, there are no long-term stability requirements an input file needs to meet beyond its creation either.
I hate AI slop and do not want it on the site. We've already had instances of people coming onto this site to post their shit vibe-codedprojectshere, and I don't see why we should tolerate this. TASing has always been about putting care and love and excellence into something, while AI output is pretty much diametrically opposed to those things. If your contribution to the site relies largely on lazy soulless unfiltered AI slop, I don't want to see it.
I think the proposal as written lines up pretty closely with my personal thoughts. I don't really have much to add to it in terms of regulations.
Memory wrote:
Plagiarism is unacceptable even if it's through a machine.
I feel less strongly about this with AI-aided research or Lua scripts than with creative output such as art, writing, or music, since it's more "removed" from the source. The problem with these cases in my view is that the original training data was stolen in the first place by AI orgs (and thus, AI orgs seizing control and power over those domains). Meanwhile, generating creative output allows you rip off other people's work much more directly. This is another reason why I am so opposed to low-quality AI slop in particular.
*NB: I haven't followed #10588: IgorOliveira66X's SNES Super Mario World "arbitrary code execution showcase" in 2:42:26.049 closely, but that submission also seemed to have other problems than just using ChatGPT for some things.
**NB: Not many people host their own models of course, and I am not aware of any (mainstream) models that actually do use ethically sourced training data, but it's still a theoretical possibility.
Joined: 10/12/2011
Posts: 6700
Location: The land down under.
JankRusty wrote:
It should also be known that Tool-Assisted Speedrunning is a form of art.
Two of my "art" pieces are titled as "suffering"; Myst & Ichima's Escape Rooms
Another way to say this, is that I'm saying to this statement and that I heavily disagree. Not all TASes would be considered as a form of "art".
Most of my TASes that I have produced have just been that. TASes. There's no care to what the user thinks from what I produced, and everything I did and done is to challenge the rules.
If someone called my TASes "art" I would look at them quizzically because takinga shit ain't it.
In relation to the topic I'm mostly neutral with this and don't mind restrictions,
After all my focus was showcased entirelyelsewhere when the first person to openly "glaze" AI had another.
"Clean-room" LLMs is probably what should be an end goal for someone instead of relying on shitty ass datacentres.
It's able to assist with what you need, especially if you're in a clinch, but regardless I rather have someone have a home server to do this shit.
Another example to go with Ruffle is RPCS3 where it's basically "understand what your code is doing".
This was written on the back of people not understanding shit they're contributing and sloppily submitting shit.
Based on that.
I could still see a fork of Hawk where non-Waterboxsystems are (theoretically) harder to get going (then again even adjacent has been hard) or one that doesn't get any work put into it cause it was done to make TASing Pokemon Colosseum easier and everything that came after was a bonus (no I'm not saying DolphinHawk is AI, just not maintained, and plenty of titles are non-working and it's not being worked on) can be made by people who do have an understanding on what the code is doing to shortcut the production of certain cores, and then an eventual future of chipping away at the slop it produces.
I've seen people wanting certain cores when queried, and certain updates to other Hawk related things such as interface and a functionaldebugger to Nymashock.
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account.Strong for yourself and also others.
Tasvideos is a gaming site, it's not a space to advance one's personal beliefs towards society or virtue signal .(Recurrent problem in TASvideos) If you don't like AI, don't use it or complain to your local geographical community. I'm baffled this discussion exists in the first place. Nobody here is a better person tha nanother for positioning themselvers against AI. In fact, small companies have a lot fo benefit from generative content of all kinds as many fo them don't have the funds to get text and art qualiity, not to mention automation of any kind in their systems.
Tas is a technical achievement that can be better enabled by AI and I'd like to see new TASers that start TASing because of it and old TASers as well be enhanced through the new tools AI allows for. I personally can't think of the benefits so far but it's pretty clear smarter people found ways to use it and i'm happy their work is being done easier and faster. This is a hobby which is often underappreacited, not very valued, takes a load of time and i don't wanna see anybody being discouraged of TAsing for using AI an inteligent way. We have limited time in the planet Earth,a and with that time some people are TASing instead of doing anything else with their lives, we should be appreciative of their dedication and we shouldn't stop people from making their lives easier with AI as long it isn't any kind of foul behavior( like plagiarism.... which to me isn't synonimous with AI ).
AI is your future and the faster you find a way to work with it in a healthy way the better, animosity between users because of AI usage just reminds of the old elitism that existed in this site for so long when "REAL TASING" was only done with "REAL GAMES(nintendo or some platofrmer)" with REAL ENTERTAINMENT, and we all know how unhappy people can get and how many communities we distanced ourselves from with exclusive attitudes.
What would be a healthy limit for AI usage in TASing? seems like a more reasomable question if i read correctly
I want all good TAS inside TASvideos, it's my motto.
TAS i'm interested:
Megaman series, specially the RPGs! Where is the mmbn1 all chips TAS we deserve? Where is the Command Mission TAS?
i'm slowly moving away from TASing fighting games for speed, maybe it's time to start finding some entertainment value in TASing.
Tasvideos is a gaming site, it's not a space to advance one's personal beliefs towards society or virtue signal.
Things in the world in fact are connected and don't exist in isolation.
i dont want your personal opinion to be used to exclude other members of this site that contribute just as much as you because you have the" right opinion" about a subject that is not even the object of this site, if you wanna ban an emulator for using proven stolen code fine go ahead but don't exclude more people for asking for help with a task that allows better tasing
I want all good TAS inside TASvideos, it's my motto.
TAS i'm interested:
Megaman series, specially the RPGs! Where is the mmbn1 all chips TAS we deserve? Where is the Command Mission TAS?
i'm slowly moving away from TASing fighting games for speed, maybe it's time to start finding some entertainment value in TASing.
You know, the folks I usually see pushing for the use of generative AI and LLM in everything use outside of this community generally are for the most part are grifters, clout chasers who don't want to do the work to be credible, and/or fascists. Those folks also said that NFTs were an inevitability that everyone has to get into or else get left behind. You know what normal people did? They pushed back hard against it and now only grifters left holding the bag are the ones still trying to sell someone on NFT schemes.
KusogeMan, whenever anything about moderation happens you immediately dive into arguing against them regardless of context and reasoning. If anything, your actions make you look like someone interacting with the TASVideos community in bad faith. I don't want those kinds of folks in any of the communities I'm in, so if you're offended then there's the door over there.
If you don't want that in TASVideos terms, let me put it in fighting game community terms. There's this post I remember reading years ago regarding a convicted domestic abuser that I still remember the context to, but not quite the wording. It went something like this:
"Just because you press buttons good doesn't give you a pass to be a shithead".
This extends to people coming to the defense of people unwelcome to the FGC community, and if I remember right was the reason the post was made. This is what I think about when it comes to creating and maintaining healthy communities, and maybe you should think on that as well.
You know, the folks I usually see pushing for the use of generative AI and LLM in everything use outside of this community generally are for the most part are grifters, clout chasers who don't want to do the work to be credible, and/or fascists. Those folks also said that NFTs were an inevitability that everyone has to get into or else get left behind. You know what normal people did? They pushed back hard against it and now only grifters left holding the bag are the ones still trying to sell someone on NFT schemes.
KusogeMan, whenever anything about moderation happens you immediately dive into arguing against them regardless of context and reasoning. If anything, your actions make you look like someone interacting with the TASVideos community in bad faith. I don't want those kinds of folks in any of the communities I'm in, so if you're offended then there's the door over there.
If you don't want that in TASVideos terms, let me put it in fighting game community terms. There's this post I remember reading years ago regarding a convicted domestic abuser that I still remember the context to, but not quite the wording. It went something like this:
"Just because you press buttons good doesn't give you a pass to be a shithead".
This extends to people coming to the defense of people unwelcome to the FGC community, and if I remember right was the reason the post was made. This is what I think about when it comes to creating and maintaining healthy communities, and maybe you should think on that as well.
ok so let me get this straight:
i argue in favor of AI USAGE because i don't want TASers for being afraid of improving their results with the right tools. Do you think It's fair to Igor who put a lot of work into his TAS for example?
f
i get called a grifter, NFT pusher( which i hate because it's clearly a scam), fascista and also not havin a real position just being a clout chaser.
And that's acceptable behavior? Do you really see this as "good person" behavior to descend into a cascade of accusationss because i disagree with position that are written as one liners?
AI is not a crime against humanity and you're not a good person for being against it. You''re not saving my rainforest by excluding members of the site for using AI in their TASing scripts.
I want all good TAS inside TASvideos, it's my motto.
TAS i'm interested:
Megaman series, specially the RPGs! Where is the mmbn1 all chips TAS we deserve? Where is the Command Mission TAS?
i'm slowly moving away from TASing fighting games for speed, maybe it's time to start finding some entertainment value in TASing.
I find the draft at EDIT 4 mostly alright, with the exception of one moment:
it is still possible to use LLMs for assistance in TASing through helping with analysis and code writing. It is, in theory, also possible for AI-generated assets to be used in an arbitrary code execution run. Due to these possibilities, we are treating these usages the same as any other contextual usage: We strongly discourage, but will still hesitantly allow, text generation in research and analysis
What does "still" mean in the bold phrase? You may prohibit it in the future?
I'll tell you what. If any policy hinders me with producing TASes in whichever way I want, I will leave this community. Or have a bunch of skilled Lua programmers monitoring #scripting channel on the Discord server, so I get my Lua scripts up and running, instead of pointlessly asking for help. I have plenty of stuff to learn and do. Lua skill is a low-priority for me, because it has no other use in my life.
If you want to play in eco-activists, go out of the streets and protest against OpenAI, Claude and other hyper-corporations. People use whatever they are offered. That's the corporations who can pull the plug off.
LLM processions is basically electricity consumption. Why don't you boycott Jaffar-Plus? It's a brute-forcer. It consumes electricity, like LLM datacenters, and you'll burn Gigawatts of electricity before actually achieving something useful in your TASes.
P.S: I'm happy NFTs failed. The first humanity victory in this 21st-century money-ruled world. Too bad cryptocurrency and mining still exist.
TASing is like making a film: only the best takes are shown in the premier.
https://xkcd.com/3246/
"Still" there means "nonetheless". If it was about time it'd be "for now".
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
LLM processions is basically electricity consumption. Why don't you boycott Jaffar-Plus? It's a brute-forcer. It consumes electricity, like LLM datacenters, and you'll burn Gigawatts of electricity before actually achieving something useful in your TASes.
I actually am a little curious on the resource consumption of jaffar now. Admittedly the difference is that the scale is so different compared to running massive data centers, but I'm still curious.
[16:08:10] <BenLubar> a TAS is just the limit of a segmented speedrun as the segment length approaches zero
[3:56 PM]Thursday, July 23, 2026 3:56 PM Samsara: i mean this in the most positive way possible
[3:56 PM]Thursday, July 23, 2026 3:56 PM Samsara: i feel like a lot of your interaction with tasvideos is finding excuses to watch TASes
I've seen instances of people giving game-specific "advice" generated from LLMs that's just flat-out wrong, and this would certainly be undesirable on shared pages where the false LLM information could be mixed into the rest of the correct, human-written information.
Having AI hallucinations on our wiki would be a nightmare of mine. Not exactly sure how to make sure we prevent it tho.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Joined: 11/13/2006
Posts: 2924
Location: Northern California
Dimon12321 wrote:
What does "still" mean in the bold phrase? You may prohibit it in the future?
"Still" as in "continue to". Sorry for the confusion there, that implication was unintentional. That's not to say our minds will never change, but I think for that to happen the industry would have to somehow get worse, and I already think it's at rock bottom.
Either way, once the policy goes into effect, I wouldn't want to make any changes to it without public approval.
Why don't you boycott Jaffar-Plus? It's a brute-forcer. It consumes electricity, like LLM datacenters, and you'll burn Gigawatts of electricity before actually achieving something useful in your TASes.
GenAI and brute forcers like Jaffar are not directly comparable. They have completely different methodologies, they serve different purposes, they have vastly different levels of outreach. I don't have to go through my Windows settings after every update to disable all the new forced-on-by-default Jaffar settings. Jaffar doesn't illegally scrape userdata to function. Hundreds of millions of people aren't using Jaffar every day, and the... I'd wager double digit number of people at most who do use it all know that they're using it and have chosen to use it. The biggest impact that I can see Jaffar having on anything is on eien86's wallet, and last time I checked that's not "the entire living world".
Why don't you boycott Jaffar-Plus? It's a brute-forcer. It consumes electricity, like LLM datacenters, and you'll burn Gigawatts of electricity before actually achieving something useful in your TASes.
Someone's home AC uses more electricity than someone using Jaffar-Plus.
