oh dear, what did they say now

So, you'd think this weird little story of the AI agent, the Santorini techbro startup, Uniuni, and Aftersell "upsell" ends there right? The bank stopped the charge. It all should have been over right?

Though some students like me do come into college thinking Marx is a waste of time, only to take a class with a Marx unit taught by an actual Marx scholar and then see that, oh, the caricature of these ideas I've been fed my whole life is completely bogus and they're important

Thanks Prof Zirelli!

US Astronaut Christina Koch using an experiment facility on board the ISS

Here's a picture of Christina Koch using hardware I designed the electronics for and wrote the firmware of

It was Christina Koch in that clip as well (or at least, she was one of the astronauts on the call, if not the one who corrected him)! No wonder the crew was silent after they came around from the far side.

If by deep ties you mean ‘has been a beloved open source project for two years and got acquired last month’ yeah. Not sure how that rounds off to ‘nazi package manager’

Wonder if anyone has made a Conway’s game of life spinner: there’s a few stationary looping automata that could be neat

You guys have that too? I thought it was just software devs.

Jesus christ this is still going on?

Title image of the comic "The music of the Spheres" by Dr. Terence Tao and Zach Weinersmith.

First 2 panels of the comic "The music of the Spheres." A man is narrating the comic, he is the comic representation of the quoted author, Dr. Terence Tao. He says "When I tell people at parties that I'm a mathematician, there's often an awkward pause in the conversation, or the obligatory..." A tall character next to him chimes in "Oh, I was sooooo bad at Math in school!" In the second panel, the narrator continues "though occasionally, you run into a math and science enthusiast who wants to discuss a recent development they heard in the news..." The narrator looks deeply touched, as a young woman is talking to him, she is saying : "oh my gosh, my friends and I cannot stop talking about reconstructing cryptographic protocols from quantum principles and... Do you want to move this conversation to a whiteboard?"

Panels 3 to 6 of the comic "The music of the Spheres." The narrator is putting a reassuring hand on a sad man's shoulder. The narration reads "...or someone who actually had a good experience in their math classes and was somewhat wistful about not pursuing it." The saddened man says "I'll always think of Hadwiger's graph theory conjecture as... the one who got away." In the next panels, the narrator states "those are good moments. But, in general, the public have very little idea what mathematics is, and what mathematicians actually do. Popular culture gives us some caricature portrayals of mathematicians"

Panels 7 and 8 of the comic "The music of the Spheres." Both panels show a caricature of a mathematicians from pop culture. The first is a tortured genius, represented by a woman holding a piece of chalk saying "I cannot know love, but by god I *will* know if every polyomino with an area divisible by six can tile and hexagon!" The narration specifies that "torturedness rates among mathematicians are, in fact, average." The second example is the savant with strange mental powers, represented by a man surrounded by equations that obscure his vision. The narration specifies that "if you see equations floating before your eyes, please consult an ophthalmologist."

Does this change my imaginary Erdos number to a complex number?

Click the link to read the full comic on the site!

COMIC ◆ www.smbc-comics.com/comic/sphere...
PATREON ◆ www.patreon.com/ZachWeinersm...
STORE ◆ smbc-store.myshopify.com

Guéguen is now up to 12 retractions and 15 "eXpReSsIoNs Of CoNcErN" at 13 different journals. docs.google.com/spreadsheets...

cc @harrisonmanley.bsky.social @jamesheathers.bsky.social

EXPRESSION OF CONCERN: 10.1111/j.1530-0277.2011.01711.x This Expression of Concern has been issued by agreement between journal Editor-in-Chief, Michael F. Miles; the Research Society on Alcohol; & Wiley Periodicals LLC. A third party raised concerns that the methodology for the study reported in this article was highly dubious given the sample size of participants. Further, the third party reported that the Materials and Methods section of the article did not include adequate information to inform how interviewers were organized, how safety was maintained for interviewers, how informed consent and sanitary measures were maintained for participants, & how errors and misleading results were minimized. Lastly, the third party raised concerns that the authors did not report funding information for the study even though the study as described in this article would require significant investment for the purchase of testing materials. The author responded to an inquiry by the publisher and stated that the original data as well as ethics approval and informed consent documentation were no longer available, and that all testing materials were donated by a relative. The author stated that no external or institutional funding was secured. Additionally, the publisher made several attempts to contact the author's institution to ask for further investigation into the research reported in this article, but no reply was received. All parties have determined that the methodology for the article does not adequately address the complex logistical, financial, and ethical requirements for a study like the one described in the article. However, because the original data are no longer available, the journal is unable to make a determination about the statistical validity and conclusions reported in the article. The Expression of Concern has been agreed to in order to inform and alert readers about the results of the investigation. The author was informed about the Expression of Concern.

Utterly pathetic response from Wiley. An "eXpReSsIoN oF cOnCeRn" that will likely hang around forever on an impossible study by Nicolas Guéguen. We are invited to believe that "a relative" sprung for the €1500–4000 worth of equipment that would have been needed. doi.org/10.1111/acer...

Didn't even finish the video before buying and installing. This looks unbelievably cool!

Fuck off! You've never seen anything like this before in your LIFE!

One `.clone()` After Another (2025)

@fasterthanli.me is this anything?

'5-node flowchart for a brain'

Terrific insult, will steal

8080

I do not wish to profit off stolen valor but Claire Wilmot’s article here - about Titus Andronicus, Jeffrey Epstein, and the terrifying efficacy of discrediting accusers - is excellent

I want to note that Ackman is now also the guy funding all of notorious fraud Francesca Gino's failing lawsuits against Harvard... seemingly for no other reason than to spite them? So mayhaps I sense more than an ounce of projection here.

Developing a streaming JSON query library for Rust+Python based on Micah Kepe's `jsongrep`: nrposner.com/blog/introdu...

Again, not a hater of the product! I use it myself, just spent a week using it to make a tool I’m happy with, and I’m finally feeling like I’ve gotten a good workflow with it. But I’m not persuaded to go whole hog on vibe coding and never reading the code it generates.

I don’t know if that translates to a criticism of CC as a product. Goodness knows there’s plenty of good products with ugly code and vv. But as a vision for the future of software engineering, yeah, it doesn’t inspire.

… with a tool they don’t trust or enjoy.

So now that we have a rare look at the internals of that product, I don’t think it’s cope for people to point out that the best tool in the sector, produced using itself, pitched as the inevitable next step of software… is held together with tape and hope?

But this isn’t just a software product like any other. It’s specifically pitched as a massive transformation of the craft at minimum, a total replacement at most. it understandable that people won’t enjoy being told that they’re obsolete and will get left behind if they don’t jump on board…

But in this context, we mostly care about coding agents. Claude Code this January was the first time I found one of these tools worth using at the price point. I haven’t tried Codex or Antigravity, but people seem to agree it’s the best or near the top

If we expand beyond the frontier labs themselves, AI companies overwhelmingly put out bad, disposable, ill-considered products, assuming they’re not outright fraudulent. Delve or the rabbit r1 are more representative of the industry than CC

"AI" is bad UX teapot from the cover of Don Norman’s “The Design of Everyday Things” clumsily ‘shopped by me "AI" means bad UX There is an emergent strain of thought in...

But that doesn’t immediately translate into good products, in terms of idea or execution. Sora was a terrible product from every angle. AI thirst trap Grok is a terrible product. The chat interface itself is arguably terrible UX inherently: buttondown.com/apperceptive...

Clearly about CC, but I’d say it’s the exception rather than the rule.

If we’re just talking about frontier AI labs, the underlying technology is generally amazing. Even the worst labs with the weakest models can do things that would have been magical five years ago.

I’m always struck by the immediate difference in atmosphere when I fly back to Spain vs US. I think the clearest was that the immigration officers in spain crack jokes as they direct you to the right line. The Americans just look dour, can’t imagine the same.