The Seattle Worldcon 2025’s WSFS Division Head Cassidy, Hugo Administrator Nicholas Whyte, and Deputy Hugo Administrator Esther MacCallum-Stewart today announced their resignations from the committee in the following statement:
Effective immediately, Cassidy (WSFS DH), Nicholas Whyte (Hugo Administrator) and Esther MacCallum-Stewart (Deputy Hugo Administrator) resign from their respective roles from the Seattle 2025 Worldcon. We do not see a path forward that enables us to make further contributions at this stage.
We want to reaffirm that no LLMs or generative AI have been used in the Hugo Awards process at any stage. Our nomination software NomNom is well-documented on GitHub for anyone to be able to review. We firmly believe in transparency for the awards process and for the Finalists who have been nominated. We believe that the Hugo Awards exist to celebrate our community which is filled with artists, authors, and fans who adore the works of our creative SFF community. Our belief in the mission of the Hugo Awards, and Worldcon in general has guided our actions in the administration of these awards, and now guides our actions in leaving the Seattle Worldcon.
Cassidy
Nicholas Whyte
Esther MacCallum-Stewart
The Seattle Worldcon’s WSFS Division administers the Hugo Awards, Business Meeting, and Site Selection. The committee’s remaining WSFS Division leadership includes Deputy Division Heads Kathryn Duval and Rosemary Parks (who is also Site Selection Coordinator).
Once before Nicholas Whyte was part of a group resignation from a Worldcon WSFS Division, in June 2021 when he was DisCon III’s WSFS Division Head (see “Another DisCon III Hugo Administration Team Resigns”).
See additional coverage here: “Responding to Controversy, Seattle Worldcon Defends Using ChatGPT to Vet Program Participants”, “Seattle 2025 Chair Apologizes for Use of ChatGPT to Vet Program Participants”, “Seattle Worldcon 2025 ChatGPT Controversy Roundup”, and “Seattle Worldcon 2025 Cancels WSFS Business Meeting Town Hall 1”.
Discover more from File 770
Subscribe to get the latest posts sent to your email.
Great. Another mortal sin committed, another skull on the wall.
A little bit of grace would do the community some good.
I think you’re misreading what’s happened. None of the people who resigned were involved in the AI dispute – two of them administer the Awards themselves, and the third is the deputy head for the Society as a whole. They weren’t under conversation.
This is reading like a protest resignation, which is much more concerning.
Wheels seem to be coming off the bus…
@:Frivyeti: Yes, that’s how I read the situation.
I haven’t followed any Seattle issues, except I gather that LLM AI was used in some way to vet program participants. Do I understand that right? Were these offices said to have been connected to AI use as well, or are these protest resignations, or what?
Ah–yes, I did misread it. And yes, that is worse.
These purity spirals are going to kill the organization. As is apparent to anyone–literally anyone–observing the situation from the outside.
I hope the concom is not imploding, exploding, or any other -ploding. The Hugos and the WSFS Business Meeting are two Worldcon departments of paramount importance. If those departments have lost their leadership even though they have not been involved in the AI controversy, this is alarming.
Because Worldcon has always, with one notable exception, taken place every year it was scheduled, I am still planning to attend. After all, it took nothing less than a world war to force the delay of PacifiCon 1 from 1942 to 1946. I hope the concom will be able to reorganize themselves by mid-August–preferably without the dubious assistance of AI dull tools.
Holy CATS.
@Andrew G, have you seen any evidence that this is a “purity spiral”? I can think of a hundred reasons for the Hugo admins to quit in five minutes.
This is…not good.
Nicholas Whyte has done more for the Hugos than anyone else I could think of during the past few decades–and I’m speaking as a four-time Hugo administrator.
If Nicholas, Esther and Cassidy have resigned, I’m confident that there was a good reason for their action.
To me it sounds like they resigned because of decisions made above them, or that they voted against and lost.
Of course Some White Guy makes the leap to “It’s purity” as if all dissent is somehow dirty and shameful.
If you are in an organization that does things you don’t agree with, you can improve from the inside, or leave. These three people chose to leave an, afaik, volunteer position requiring significant work, in an organization making decisions that are unpalatable for its stakeholders/constituents.
Given our very strong grapevine, I think we’ll all know why in 72 hours or less. Rather than baselessly speculate, I’ll just wait. You do you, boo.
I don’t quite get why Nicholas and his colleagues are jumping ship, other than a misfired display of virtue-signalling which will actively undermine the members’ convention experience by removing the team’s undoubted expertise and integrity from play. They did nothing wrong, and no one believes they did. A solid lose-lose.
You mean aside from leaving an organization that almost certainly violated the GDPR by feeding unredacted PII into chatGPT during thier “vetting process”?
As an American, I know we’re used to our data being sloshed around like water in a bath tub, but the rest of the world isn’t nearly so flippant.
Additionally, the way the con chair describes thier use of an llm seems to indicate gross incompetence, at best.
Jeepers. I think this may well signal the end of WorldCon.
But kudos to them: they are some of the best people in fandom and this is what people should have done in Chengdu.
@Anna: No, this isn’t a signal of the end-times for Worldcon. Post-2026, there are healthy bids from Montreal (2027), Brisbane (2028), Dublin (2029), and Edmonton (2030). The 2026 Worldcon in Anaheim may well be the last US Worldcon, however.
I’m sure Nicholas would have explained why, if he could ethically.
I think the next move is up to the Chair. I hope it is something that can be resolved constructively. Otherwise it is going to be hard to find anyone who is both competent and willing to take on the job.
I wrote: “I don’t quite get why Nicholas and his colleagues are jumping ship, other than a misfired display of virtue-signalling which will actively undermine the members’ convention experience by removing the team’s undoubted expertise and integrity from play. They did nothing wrong, and no one believes they did. A solid lose-lose.”
Greg replied: “You mean aside from leaving an organization that almost certainly violated the GDPR by feeding unredacted PII into chatGPT during thier (sic) ‘vetting process’?”
Big f***ing deal. None of that reflects badly in any sense upon the Hugo Awards team, and last time I checked (having chaired three Novacons and worked on numerous programme teams), wasn’t the entire idea about creating a great experience for the convention members? Or is personal grandstanding our be-all and end-all now?
Is there some code/contract that prevents people from disclosing questionable behavior at this level of committee administration? If there’s a problem, why can’t they just disclose it?
Tom Becker replied: “I’m sure Nicholas would have explained why, if he could ethically.”
I wish he’d at least try. From my own PoV, this seems a mis-step for a team which has previously been an exemplar for ethical transparency.
Jeanne Jackson wrote: “The 2026 Worldcon in Anaheim may well be the last US Worldcon, however.”
Seriously? Please show your working towards reaching that conclusion.
@TMTMTMTM — no, nobody has accused any of the people involved of using LLMs in any way. The LLM use was entirely based around vetting panel participants, not anything to do with the Hugos. But presumably the three people who’ve resigned felt it would be unethical to continue to work with an organisation that had allowed that (and I can’t blame them).
@Steve Green — the GDPR situation might not “reflect badly” on the Hugo administrators, but might, for example, leave them legally liable for misuse of data.
As for “creating a great experience for the convention members” they might reasonably think that that is an impossible aim working with people who have shown a willingness to discard ethics and disregard privacy and data handling laws related to those convention members, and that this violation is so egregious that it is simply impossible to run a convention with that aim.
It’s not “virtue signalling” to refuse to do a job you believe cannot be done, or to refuse to work with people you think are behaving unethically, incompetently, or quite possibly illegally.
@Steve Green: OK, I’m being pessimistic. It’s possible that, after the Trump regime ends, we might have a chance to clean up the mess he has been making–starting with sharply curtailing the Gestapo (ICE and FBI). The underlying problem, though, is the intense xenophobia which made Trump possible, and which has made many foreigners choose to avoid travel to the US. I don’t think international fandom will want to return Worldcon to the US until we can demonstrate that we, as a nation, have returned to being hospitable, rather than hostile, to foreign visitors–and can be trusted to remain so, no matter who is in high office.
@Colin Kuskie: it might not be outright prohibited, but it could still be legally imprudent to make any statement about the activities of third parties that might have been involved without running it by a lawyer first, transparency or no. I believe Worldcon 76 wound up paying out a pretty sizeable amount to settle a defamation case because of slightly infelicitous wording in the Con’s good faith attempt at explaining why they had refused to allow a certain member to attend and refunded the membership. See the section on “Privilege” here, where the Con was arguing that the whole membership needed to know that information in the interests of transparency, and the Court was all “not necessarily, that question will need to be decided by a jury.” (much more expensive than getting it resolved at the motion to dismiss stage).
The final paragraph of Rick Moen’s comment there seems apt:
I see no evidence, other than the coincidence of time, that this is related to the programming vetting process teacup tempest. In fact, they seem to go out of their way to say no LLMs were involved in their tasks. Something else, but no transparency about it, sadly unsurprisingly.
I wrote a very long tractatus here and then erased it. Worldcon is showing signs of essential failure.
This isn’t the first time Worldcons have had multiple resignations.
This isn’t the first time there’s been a management scandal–if that’s what’s happening here.
@Colin Kuskie & Kate–If this is in any way a personnel issue, yes, there’s a reason to keep silent about it, and no, we won’t know in 72 hours what’s behind this. Or likely ever.
That’s because releasing any information about it would make the Seattle Worldcon and possibly individual officers of it, vulnerable to lawsuits. You may recall that one Worldcon, not all that long ago, made that mistake, in the service of transparency, and got sued–and had to litigate that lawsuit. It was a mistake, and a mistake experienced experienced conrunners aren’t going to repeat. It’s bad governance. (Yes, that’s corporate speak. Each individual Worldcon is a non-profit corporation in the US, and I believe in most countries that have hosted Worldcons. Even if I’m mistaken about that last bit, it doesn’t matter. This one is in the US, governed by US law.
If the reason for this People Did Something Awful, Seattle and its officers will not be saying anything that could damage anyone’s reputation, unless that person or persons go public first.
As for this being the end of Worldcon, also no. LA is a completely separate organization, and will if necessary be warned privately about people who should not be holding responsible positions.
After LA, the next several years are already going to be outside the US.
What’s happening is disturbing and a problem. But that’s hardly new for Worldcon, going back to the first Worldcon and the Exclusion Act.
Side question: Did any of the Chengdu honorees ever get their Hugos?
@Madame Hardy: Chris Barkley contacted all the U.S.-based Hugo winners from 2023 that he could reach, and posted about whether they received their awards. See here.
Some of them received their Hugos intact.
Others received Hugo trophies that were damaged.
At least one winner took her own Hugo trophy home from Chengdu herself rather than having it shipped.
Some others didn’t receive their trophies at all.
Two of them refused their Hugos altogether.
And still others didn’t respond to Chris’s inquiries or he couldn’t reach them.
@Steve Green: I see two possibilities.
One, Anaheim is the last US Worldcon.
Two, Anaheim still takes place, but by the time it does California is no longer part of the US, making Seattle the last US Worldcon.
Reading between the lines, the fact that the use of LLMs (or rather mansplain generators) were mentioned first thing in the second paragraph is a pretty big hint about the substance of the dispute.
I think the resignation also is the reason for the cancelled town hall. The WSFS DH would necessarily have been extremely involved in that process.
Karl-Johan: Your thought about the reason for the cancelation of the Town Hall is intriguing.
If Seattle had one team that spent many hours doing tasks manually, and another that saved many hours by using AI, I can see the tensions between the purist and pragmatic approaches. But if that means that convention committees need to publicly commit to being either AI inclusive or inclusive, that will reduce the number of volunteers, either doing the thankless tasks, or not wanting to be attacked for finding ways of speeding up those tasks.
Either way, you will just stop getting volunteers. I long ago decided to steer clear of con-running for all the burdens fandom now want to impose on them.
I thought it was pretty clearly implied that this is due to a moral objection to the use of LLM/genAI on the basis that it went against the spirit of “[celebrating] our community which is filled with artists, authors, and fans who adore the works of our creative SFF community”, as those computer models are built off the backs of stealing that work, both professional and fannish.
I could, of course, be completely wrong and reading too much into it.
Don’t know what is going on here.
How could one use LLM AIs in counting Hugo nominations???
Why the public resignation?
Is this virtue signalling?
Yes, Seattle got it wrong regarding AI use in vetting programme participants, but didn’t they do the right thing when it was pointed out to them and they apologised and rolled back on their decision?
Anyway, zero confidence in Esther MacCallum-Stewart. The Glasgow con broke the WSFS constitution and rules of continuing effect. (As someone who dug into their pocket and supported Glasgow from bid stage as a super friend member who was unable to go and got zilch she has broken my trust in current generation of worldcon SMOFs – I’ll never pre-support or supporting member a Worldcon again!)
1.5.2. Constitution
https://www.wsfs.org/wp-content/uploads/2024/10/WSFS-Constitution-as-of-20240812.pdf
Rules of Continuing effect voted on at Helsinki Worldcon BM-2017-01
https://www.wsfs.org/wp-content/uploads/2024/07/WSFS-BM-Resolutions-Rulings-of-Continuing-Effect-20240706.pdf
John Bray: There’s nothing ‘practical’ about LLM use here. From their own description the use they made of it was one an LLM is useless at; asking it to search someone’s background and provide links to anything problematic. LLMs cannot, by their nature, make judgement calls like that. They are a predictive engine that simply provides what their training data suggests is statistically likely to be the next word.
They so regularly hallucinate sources that multiple lawyers have been sanctioned for relying on them in briefs where they have invented precedents and case references that are non-existent.
Using one for this purpose is base foolishness.
Ah yes, the dreaded virtue signalling problem, which seems to be the way people now decry doing the right thing. For example, say you are holding a convention in a country without freedom of speech – I mean a country other than the US, of course. And say that the concom asks you to vet the Hugo candidates for their views and general suitability to be honored with an award, so that they can be deemed ineligible.
Now the Hugo administration could have done the honorable thing and resigned en masse. Or they could have shrugged, said, eh, when in Rome, and carried on purging the ranks of people with unacceptable views.
The first course of action being labelled as “virtue signalling” is why the Chengu Hugos were a travesty.
It is a bit like being accused of being social justice warriors: because obviously social justice is not something worthy of being fought for.
As for the end of WorldCon, I have seen too many instances of the wider geopolitical environment making each location a possible hornets nest – there are not now, but there could being the future problems providing toilets of attendants at UK cons, for example.
Is it possible to have a WorldCon in a world where country after country becomes unsafe? Is it possible to have a WorldCon in a world where the country contributing the majority of attendees is not longer safe?
Just like the first time somebody refuses to obey a Supreme Court pronouncement and says “let them enforce it”, there are times when the Hugo finalists list is purged on ideological grounds, and the rule once being broken, the silent covenant that underpins the rule of law being rejected, there is no going back.
Bummer to see these resignations. Nicholas Whyte is a great Hugo administrator.
Thanks to them and the rest of the Hugo volunteers for their efforts. Hope this is the last turmoil for Seattle Worldcon but it probably isn’t. Are we having fun yet?
I’m seriously bummed. I always feel a little more confident in the Hugo Awards when Nicholas Whyte is administering them.
I don’t think it will be possible to have a Worldcon in the USA for a while simply because it’s foolish to come there right now. You’ll have to settle for a USAcon. Our governments outside the USA are warning against visiting USA right now as trying to get into the country you have a big chance of being denied entry, and having the wrong skin colour might get you not returned home, but potentially sent to the concentration camp in El salvador. USA needs to fix its dictator problem if it’s to host another Worldcon.
Mm.. very saddened to see these three prominent Worldcon people resign-especially Nicholas. Originally I had sympathy with 2025/Seattle Worldcon: it was voted in, pre the arrival in The White House of and whom I call (and no doubt now I’m a marked man by ICE, if I ever visit the US of A) “the orange tinted mega-narcissist”. And it (along with 2026/LA-Anaheim) now have a situation which they didn’t envisage -re actual attendees from overseas.. But does Seattle (re Chat GPT) have some sort of death wish?? (I’m not attending either Seattle or LA anyway…) Best wishes …
I had a go with Gemini LLM with prompts “are the views of X compatible with the Seattle 2025 Worldcon’s code of conduct”.. I tried it with wrong-uns, good-uns, and wrong-uns-for-fannish-actions and each time it worked well providing a summary and references. I didn’t follow up the references or check for edge-cases, because I don’t care that much about programme participant vetting. If my boss had asked me to do what I thought was a pointless task, I’d delight in getting the LLM do rattle through it, as I would writing marketing puff or risk assessments.
MODERATOR’S NOTE: If someone doesn’t respond to my attempt to authenticate their registration email address, I don’t approve their comment. This impacts a person trying to leave their first comment.
@ Kate, no, not that one
Of course. The only form of bigotry that may be practiced openly and without penalty.
Back to the main topic, I appreciate Nicholas’ work and find it regrettable that he was in a situation where this was the best option.
Regards,
Dann
People sleep peaceably in their beds at night only because rough men stand ready to do violence on their behalf. – Richard Grenier
@John Bray–If you didn’t follow up on the references Gemini LLM provided, you only know that it was easy, and that the references looked plausible. You have no clue whether those references are real, or completely fabricated.
Lawyers using LLMs to get cites to use in their briefs have had their licenses suspended or been disbarred, because the very plausible legal cites turned out to be completely fictional.
This is not a good use of LLMs.
The LLMs of a few years back might invent references but the modern ones have incorporated web searches into their discovery logic. It would be trivial for them to only provide URLs and references that were valid data points inside current search engine space. They might still come up with spurious arguments, but they would not be backed up by references. It was such a public complaint that the developers will have fixed the loophole.
I asked Gemini “write a social media post that expresses outrage about the 2025 Worldcon using AI to vet programme participants against its code of conduct” and it returned a great topical rant.
@John Bray,
Conservative pundit and sci-fi/fantasy fan Jonah Goldberg has been testing various AI platforms by asking it questions about himself and about topics with which he is intimately familiar. None of those platforms have been able to provide responses that are remotely accurate.
Obviously this is anec-data. But Jonah is enough of a techno-geek and a journalist that his diligence in this area can be presumed to justify his conclusions.
Regards,
Dann
“You can’t be a real country unless you have a beer and an airline. It helps if you have some kind of a football team, or some nuclear weapons, but at the very least you need a beer.” – Frank Zappa
@Anna:
Ouch. Quite right.
What they should have done:
1. ask for volunteers. openly, for everyone. I’d love to help but I can’t pay for a ticket
2. If not enough volunteers, make an announcement: “we will use AI if we don’t find x amount of volunteers for y date bc we don’t have enough people”. volunteers would appear.
3. IF still not enough volunteers, ANNOUNCE that you WILL use AI and say exactly how and which and allow any applicants who want their application reviewed manually to opt out
4. profit! or at least no drama.
Andrew Gillsmith “a little bit of grace”??? But not towards the POC that were unfairly rejected by the racist plagiarism machine??
@W. Maxwell
“From their own description the use they made of it was one an LLM is useless at; asking it to search someone’s background and provide links to anything problematic. ”
I don’t think it is useless. I asked Google Gemini “Is there anything controversial in Dave Mccarty’s background?” and it responded
“Based on the search results, there was a controversy surrounding Dave McCarty in early 2024 related to the Hugo Awards. He was the administrator for the awards and was censured and resigned after a public outcry regarding a document that showed certain nominated works were deemed “not eligible.” . . . Additionally, in April 2025, there was a report of a legal issue where Chris Barkley was suing Dave McCarty in small claims court to obtain his 2023 Hugo Award trophy or $3000.” It also provided links to sources that confirmed these statements (https://lunch.publishersmarketplace.com/ and https://file770.com/).
If I Google [“dave mccarty” controversy], I get assorted links documenting the same issues (reddit, Publisher’s Weekly, Camestros Felaptron’s blog, etc. However, it takes a moderate amount of time to go through these links to get a succinct statement of what the actual problem is, such that a decision can be made independently by a real person “should Dave McCarty be included in a WSFS program?”
I repeated the test with Jon del Arroz, and got similar results: While both the “AI’ and the search engine gave useful data for the problem at hand (identifying issues that may mark someone as a bad candidate for a Worldcon panel), the “AI” did it in a way that is faster and more direct and useful for my purposes.
So, one of the criticisms of using the AI models (“They don’t even work!”) is not true.
Another criticism is that they “steal” the work of real people. Is there any substantive way in which an LLM’s use of copyrighted texts (in general) is different from a search engine? I’ve made many Google searches over the years for which one of the results was an excerpt from a copyrighted book archived at Google Books, or a link to a news article behind a paywall. If I Google “I could have become a mass murderer after I hacked my governor module” (from the first sentence of the first Murderbot novel), I get many results that all point to Wells’ copyrighted work. Is there any way that Google the search engine could have known thiese things without processing data in pretty much the same way that an LLM does? Google has scanned/scraped thousands of copyrighted books in Google Books, and millions of copyrighted news articles for Google News. LLMs are not inherently different from the major search engines in this respect.
But what about “AI’s are resource hogs, using electricity and water”?
Again, search engines carry the same burden. See “Cooling the servers, without breaking the bank” (NY Times, 11/7/2007); “Renewable Energy Can’t Run the Cloud; Data centers now consume about 1.3% of all global electricity” (Wall Street Journal, 5/28/2012).
You can vet people for panels, or you can avoid using systems that use people’s work and intensively use electricity and water. You can’t do both.