Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
Due to their behavior surrounding this submission...
Submitting another author's work without their permission.
Taking credit for work that was not theirs.
Posting ROM links several times.
Taking umbrage with Asumeh attempting to make the run verifiable, hypocritically by demanding "permission" they themselves did not ask for when submitting the run.
...duckthegun0 has been indefinitely banned from TASVideos.
BLGHA, you're free to submit your own work here on your own terms.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
The policy has been put into place. Previous revisions of the second draft remain available in the page history. The original draft and its revisions are in the process of being backed up elsewhere. Thank you to everyone who pitched in and helped refine the policy.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
In response to the rapid proliferation of generative AI across the tech sphere and the numerous environmental, ethical, and mental health concerns that are arising from it, we as TASVideos staff have put out a statement describing our stance and the actions the site will be taking against generative AI use going forward: This page explains our stance and the rules that will apply to all TASVideos users going forward.
This page has been drafted and discussed amongst staff and community members for the past several weeks and is unlikely to change further, apart from minor changes to wording and clarity. For more information on how this policy came into place, please see this discussion thread.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
Thanks for the quick response. I've gone ahead and removed the links in accordance with the upcoming policy. The disclosure is appreciated, of course: I've left it in case anyone's interested and wishes to contact you directly about it.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
Sorry, what's the point of all the AI agent stuff in the submission text? Is any of it actually relevant to the submission? What are Judges supposed to glean from it?
...Also, are you aware that we've been discussing banning that sort of AI usage in submission texts and across the site?
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
Alyosha wrote:
So, would using such an emulator be acceptable for submissions? Do we care about the provenance of AI generated code? Under the current proposal it is my interpretation that this emulator would be approved.
I would say no, here. Admittedly the draft isn't quite clear about what we would do with new emulators, but I think the spirit of the law is that we wouldn't actively add support for anything with AI-generated code at this time. Exceptions might be made if there are no other alternatives, i.e consoles that we don't already support in some fashion, but for something already widely supported like NES I see absolutely no reason to add some brand new vibecoded alternative, especially when we can add other options that aren't.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
Spikestuff wrote:
Since the intent was "what about on the Hawk side of things"
Alright, so I completely missed this at every possible opportunity.
Spike and I talked about it in private, but just to provide a public response: TASVideos and BizHawk are technically separate entities. Hawk started here and is still primarily dev'd by TASVideos members, but it's definitely its own thing and isn't covered by the policy. I would like to see BizHawk devs come to their own conclusion about AI-assistance, but that's outside of the scope of this thread and this policy as a whole.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
I'm having a lot of trouble figuring out what the proposal actually is, here. Nothing jumps out to me as an actual suggestion. If I'm understanding it correctly it has to do with local environments, and the short answer there is that I don't want to make any distinctions there. Local or not, it's the same tech with a lot of the same concerns, just that opening the door to local environments introduces some major, difficult to moderate loopholes without actually adding any real benefits to good faith users.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
YoshiRulz wrote:
AI ban is still not checked in to the repo.
Presumably it will be when this becomes official.
Would you consider allowing AI-assisted decomp/disasm in userfiles (Lua scripts) and wiki pages? Also, are homepages included in wiki pages?
As the rest of the policy should have made clear, anything directly containing AI-generated content is not allowed. If it doesn't directly contain anything AI-generated, it's fine. Yes, homepages are included, because they are wiki pages.
??? LLMs are neural networks.
"Non-agentic" is meant to imply anything outside of the current prompt-based plagiarism agents. No ChatGPT, no Claude, etc. If it's that unclear, then at least give me a suggestion on how to word it properly instead of trying to educate me on something I have no interest in learning further.
If you have a published run that you would prefer an alternate metric for, or if you know of a published run that has requested such in the past, post it here and we'll change it as soon as we're able to.
Ah, I see this is literal. The first ~11300 frames are identical. In the future, if you're directly using input from other people, please credit them by adding them as co-authors. I think in this case, however...
lord_tompa
...might not want to be associated with this submission for more than just the misnaming.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
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".
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
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!
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
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.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
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!
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
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:UPDATE (August 1st, 2026): Brand new, fully rewritten draft just dropped. For comparison purposes, the original is still located here.UPDATE (August 11st, 2026):The rewritten draft, after a few further revisions, is now live as the official site stance and policy.
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.
Published Author, Site Admin, Expert player
(2350)
Location: Northern California
Joined: 11/13/2006
Posts: 2938
Location: Northern California
NxCy wrote:
Does the site have an official stance/policy on LLM usage in preparing TASes? I had a quick search, but I'm not really sure where to look. This thread gives the impression that any mention of an LLM makes a submission unacceptable.
So, the timing of this submission is wild, because we've actually been working on one since late last month. We've been refining it slowly since then, but this submission and the reaction to it prompted us to pick up the pace and make it more of a priority.
It's still in the refinement phase but once all of us as staff are happy with it I'll be bringing it to the community for further feedback.
