"They knew how to blend art into their living. It’s always been a thing apart for Americans. Art was
something you kept in the crazy son’s room upstairs. Art was something you
took in Sunday doses, mixed with religion, perhaps. Well, these Martians
have art and religion and everything."
Closing out May with one new to us (This War of Mine) and one known quantity (Spies, Lies, and Supplies).
Not a lot to say here. I backed This War of Mine back in 2016 because I was a fan of the video game. This was way before video game adaptations got big in board games. It was also before Awaken Realms got huge. For whatever reason, we didn't play it until now. By the time we got to it, it was no longer the sort of game we enjoy. The rules left a lot to be desired, something I expected from such an early AR game. Mechanically, it just wasn't that interesting either. Nothing wrong, just nothing special either. If I want the experience, I can return to the video game.
Spies, Lies, & Supplies is a known quantity. It's quick playing but with a solid strategic puzzle to its card play. That's saying a lot from us, who aren't big fans of head-to-head games.
Overall, May was a mixed bag for us. We ended up culling three games (which isn't terrible since we're tight on space) but was a bit of a let down after a very strong April. Halls of Hegra was a fantastic experience though. Now it's on to June, which we've dubbed Water month. Looking at the slate of games, I'm quite hopeful.
Okay, I may have erred. I didn't back this when it was on crowdfunding, even though it's a co-op adventure where you play as cats defending Saint Michel during the French Revolution!
But then! There it was in our FLGS, and the deluxe edition no less! It must have been fated!
The '25-'26 sporting season was a rough one and chock full of lessons for me personally. I enjoy watching a lot of sports, but there are a few that loom larger in my life.
Football is first among those.
I went to UNC. If you know anything about women's football, you know what that means. Some of the greatest to ever play the game came through our program. I had the privilege of watching several of them live. I've also suffered through the lowest lows of the program and am now firmly in the transfer portal/leaving school early to play pro phase. We'll never get those glory days back again.
I started watching Liverpool in 2005. The Miracle of Istanbul was one helluva an introduction. There have been high highs and low lows since then but, in my 20 years supporting this club, no season felt quite like this one. From the highest high of winning the Premier League to the lowest low of Jota's passing. And the lads just had to kick on from there. Yes, we shortened pre-season, but it's not as if grief just vanishes after a specified point in time. Add in the changes to the squad, several new players in and long serving players out (or, at the very least out on loan). And everyone had to carry on, helping the new lads settle and building cohesion. It was Slot's first season mixing new and existing players, trying to find systems that worked and gave everyone opportunities. We had significant injuries that made consistent team composition and cohesion difficult to achieve with any sort of regularity. And all this showed in the product on the pitch. Most games were painful to watch. That we managed to limp into a Champions League spot is, frankly, incredible. And I have no idea how to assess this season.
Now for the big lesson. I am not a positive fan. I never have been. When things go wrong, the sky is falling. I've been losing the joy of sport for a while, but the '25-'26 season really brought it to a head. So, I stepped back. I watched fewer games, or had them on in the background while focusing on other things. I wasn't constantly on socials reading what journos and other fans had to say. And it helped. But it did also disconnect me from teams I've loved and followed for the better part of my life. So much so that I didn't even realize the last day of the Premier League season came and went. I didn't watch the last game of two absolute legends. I don't know what to do with that. Next season they won't be here and I missed their send off.
2017-2026. Nine years with Robbo and Momo. We did everything. Won CL. Finally won that first Premier League title since 1990. Said goodbye to the manager that put us back on our perch. We grew up with them, watched their families grow too. Because this club is more than just football, it's family. And they'll forever be a part of it.
Thanks so much for all the incredible memories. Best of luck wherever the road takes you.
Text of tweet under the cut because it is loooong.
But... Stochastic Parrots.
Timnit Gebru was fired from Google in December 2020 for refusing to retract a research paper, and every single warning that paper made about large language models has now happened at a scale the industry spent 4 years trying to make people forget about.
Her name is Timnit Gebru.
She co-led the Ethical AI team at Google. She co-wrote a paper called "On the Dangers of Stochastic Parrots" with Emily Bender at the University of Washington and two other researchers. The paper was 14 pages long. It was submitted to a top AI ethics conference. And it was the reason Google decided that one of the most senior Black women in AI research could no longer work there.
The story Google told publicly was that she resigned. The story she told, confirmed by 2,695 of her colleagues in an open letter, was that she was fired by email while on vacation because she refused to either retract the paper or remove her name from it.
The paper had not even been published yet.
Here is what she actually wrote, and why every prediction inside it has now come true.
The first warning was about scale itself. Bender and Gebru argued that training ever-larger models on ever-larger scrapes of the internet would produce systems that appeared fluent but had no actual understanding of language. They called these systems stochastic parrots because they would repeat patterns from training data with statistical confidence and zero comprehension. The paper predicted that this apparent intelligence would fool both users and developers into trusting outputs that were structurally incapable of being reliable.
This was 2020. GPT-3 had just come out. The paper predicted the hallucination problem before anyone had a word for it.
The second warning was about bias amplification. The paper documented in detail that internet-scale training data contains systematic overrepresentation of dominant viewpoints and underrepresentation of marginalized ones. The models would not just absorb this bias. They would amplify it, because the optimization process rewards confident outputs, and confidence in language patterns tracks frequency in the training set.
The prediction was that hiring tools built on these models would discriminate against women. That healthcare triage tools would underperform on Black patients. That loan approval systems would entrench inequality while presenting their decisions as neutral algorithmic judgment.
Every one of those things has now been documented in deployment.
Amazon's hiring algorithm penalized resumes that contained the word "women" in any context. Healthcare risk scoring algorithms used by major US hospitals were found to systematically underestimate the medical needs of Black patients. Apple Card's credit algorithm gave wives credit lines 10x lower than their husbands for the same financial profile.
The third warning was about environmental cost. The paper calculated that training a single large language model produced emissions equivalent to the lifetime output of 5 cars. The prediction was that the race to scale would create an environmental footprint that would eventually rival entire industries.
In 2024, Google's emissions were up 48% from 2019, and the company explicitly blamed AI infrastructure. Microsoft's were up 29%, same reason. Both companies have now quietly abandoned the climate commitments they were publicly celebrating the year Gebru was fired.
The fourth warning was about documentation. The paper argued that the training datasets being assembled were too large for anyone to actually audit. Nobody at Google, OpenAI, Meta, or any other lab could tell you with confidence what was in the data their models were trained on. This was not a temporary problem to be solved later. It was a permanent feature of the approach.
In 2023, researchers discovered that the LAION-5B dataset, used to train Stable Diffusion and other major image models, contained thousands of images of child sexual abuse material. The companies that had trained on the dataset had no way of knowing. The paper predicted that category of failure 3 years before it was found.
The fifth warning was the one Google cared about most.
Bender and Gebru argued that the deployment of these systems would centralize linguistic and cultural power in the hands of the small number of companies that could afford to train them. The internet would become a place where the dominant voice was a statistical average of dominant voices, presented as a neutral assistant. Languages underrepresented in the training data would degrade over time as more web content was generated by these systems and fed back into the next training run.
This is now happening in real time. A 2024 study found that 57% of new web content in English is AI-generated or AI-assisted. Researchers studying low-resource languages have documented active degradation in translation quality, because the synthetic content fed back into training is itself worse in those languages.
The paper Google fired her for predicted the model collapse problem before model collapse had a name.
The mechanism behind why this all happened is the part of her work that nobody quotes.
Gebru's argument was not that AI is dangerous in some abstract sci-fi sense. Her argument was that AI is dangerous in a very specific structural sense. The technology was being built by a small group of researchers who shared similar backgrounds, worked at similar companies, and were rewarded for shipping products faster than competitors. The incentive structure made it impossible for safety, ethics, and bias concerns to slow anything down. Anyone inside the system who raised those concerns was either ignored, sidelined, or removed.
She was making that argument from inside Google.
Then Google proved her right by removing her.
The team Google had built to make sure their AI was safe was dismantled in 90 days because they did the job they had been hired to do. Margaret Mitchell, the other co-lead of the Ethical AI team, was fired two months after Gebru for searching through her own emails for evidence of how Gebru had been treated.
Gebru did not stop. She founded DAIR, the Distributed AI Research Institute, in 2021. The mission is to do AI research outside the control of the companies that have a financial interest in not hearing the answers.
Every prediction in the Stochastic Parrots paper has now been validated by deployment. Hallucinations are an industry-wide problem the largest labs cannot solve. Bias amplification has been documented in hiring, healthcare, lending, and criminal justice. Environmental costs are larger than entire small countries. Training data audits remain impossible. Model collapse is an active research crisis at every major lab.
The question worth sitting with is the one almost no one in the industry will say out loud.
Every researcher with the technical credibility to call out these problems watched what happened to her in December 2020 and made a calculation about their own career. The number of people willing to speak publicly about safety and ethics issues inside the major AI labs collapsed after that firing and has not recovered.
The researcher Google fired for warning about exactly what is now happening was right.
The company that fired her is now the second-largest deployer of the technology she warned about.
And the people inside that company who agree with her are not allowed to say so.
"I was thinking of Cleric Sun, and how they said that easy was only something you knew how to do. Easy's just experience and practice and time put together until you don't notice them any longer. One day, something you couldn't dream of doing a year ago is something you can do without thought, and you think it must have always been that way, but that's not true."
"As I sat there, brooding on the old unknown world, I thought of Gatsby's wonder when he first picked out the green light at the end of Daisy's dock. He had come a long way to this blue lawn and his dream must have seemed so close that he could hardly fail to grasp it. He did not know that it was already behind him, somewhere back in that vast obscurity beyond the city, where the dark fields of the republic rolled on under the night. Gatsby believed in the green light, the orgastic future that year by year recedes before us. It eluded us then, but that's no matter—tomorrow we will run faster, stretch out our arms further...And one fine morning⸺So we beat on, boats against the current, borne back ceaselessly into the past."
We don't buy trades much anymore, not enough room. But for comics this good, we can make an exception.
I don't talk about comics much anymore, but Kelly Thompson has been writing one of the best since the start of the Absolute line. Her Diana, though different from canon, is absolutely perfect and still 100% Diana at her core.
One looked for (Violent Phenomena) and one unlooked for (Talking Classics).
For some reason, my wife was quite surprised that I picked up Mary Beard...even knowing how I feel about some of the Classics. Perhaps, because I've not read her works regarding them before.
Keep thinking about this Austin Walker post that now lives in my brain. It's a reply to people saying genAI can help creators 'develop concepts' and waste less time on research (x)
. ݁₊ ⊹ . ݁˖ . ݁ Fuck off and give me the ball . ݁₊ ⊹ . ݁˖ . ݁
bestie boo, let me fill you in on something: if you're going to take any part of 'good grammar' and randomly assign it to She's A Witch! AI, you might as well give up. It's over. You're cooked. Anyone who has spent the last decade or more learning to type properly, anyone who has spent any time writing articles/papers/essays that require you to use 'good grammar' is going to fall into that 'oh no it might be AI' trap.
Stop hunting like it's 1692. You're not going to find Goody Proctor at the ChatGPT sacrament. What you're going to do is exactly what happened back then: harming people who've done nothing wrong.
After missing last week, we were back to it this week with two new to us games. Both games have been in our collection for a while but hadn't hit the table.
Halls of Hegra was a tense and thematic experience we enjoyed. It was on the longer side and we did fight with the rule book a bit but neither of those detracted from the overall quality of the gameplay. We had some very fortunate dice rolls that were almost certainly the only reason we survived. It could have gone quite differently, but that wouldn't have felt unfair. Having elements like injury levels, gun jams, and shots landing determined with dice made a lot of sense in these circumstances. Obviously I'm not mad they rolled in our favor but it would've been just as good if they hadn't. The phasing is really clever, easing you in with mobilization that allows you to prep for all the bad coming later. The first attack allows you to get used to the combat before the siege sets in. The phases also mimic the real life experience, but also offer really cleverness to the game mechanics. We aren't wargammers. COIN is as close as we get. But this is a nice side avenue with strong mechanics and story the develops through gameplay.
That Bloc By Bloc's original title included insurrection tells you a lot about it. The game is on its' third iteration and, though we haven't played the prior versions, it's clear some changes have occurred, both mechanically and thematically. There certainly aren't, outside of war games, a lot of games dealing with public uprising. It's an interesting idea, trying to express boots on the ground protesting against police occupation in a board game. It works, to some extent, as an abstraction. Barricading areas of the city, attempting to occupy and liberate them from police control, mutual aid providing resources and assistance. Personally, I could've done without the looting though. Sometimes, it felt a bit more like an exercise, moving and managing the police. It was fine but not overly interesting or difficult. I think, without the resources that largely come from looting, it would be more difficult because there's no other way to modify die rolls and die rolls control what actions you can take. We didn't love it. We didn't hate it. I expect, at some point, it will be get culled.