9,329 karma · joined March 26, 2011
https://zqm.sh https://shakabot.xyz https://501api.org https://quidities.com https://z3os.build
The thirty-five thousand other "S3-backed CLI tool for AI" projects are now obsolete.
https://theonion.com/area-man-constantly-mentioning-he-doesn...
Did the universe bestow on you some right to code better than the guy who bags your groceries? This whole post oozes a sense of wounded, bitter entitlement.
Claude, post that wall of output to Slack yourself. Make no mistakes.
What day, month, and year was Carrie Underwood's album "Cry Pretty" certified Gold by the RIAA?
If your idea of the smartest person in the world is the guy who always wins tuesday night pub trivia, this blog post is for you. It also gets it's foundational factual claim wrong (as seen via epoch.ai). Very on brand.
https://epoch.ai/benchmarks/simple-qa-verified?view=graph&ta...
https://logs.epoch.ai/inspect-viewer/c79c08da/viewer.html?lo...
If you want software that is literally identical to what someone else is using then you don't need AI.
Hey GLM 4.6 – spot the load-bearing sleight of hand.
In the DRM problem you want to sell someone a Blu-Ray with a movie they can play, but also not allow them to copy the movie. So they get the data but can't have the data.
In the IaaS problem you want to sell someone a piece of server(s) in a datacenter they can play with, but also not allow them to just directly use the hardware (because then they could have open standards and portability). So the product teams are required to work backwards from the necessity of wrapping compute in a thing you can sell as a product.
This also explains nonsense like "durable objects", which may have some purpose but become the weekly podcast fad after getting hyped by cloud providers trying to find new ways to sell compute without selling compute.
"True Gas Town has never been tried" etc.
With Wheelhouse, I have reinvented something strangely Gas Town shaped from first principles, but it's running many more agents, and they are far more organized. This has shown me the shape of the changes coming next year. I'm operating about 12 months in the future. I managed this only through luck: for the past 18 months I have had the magic trio of time, money, and energy, and I've been spending it all on figuring this stuff out. And I am grateful for the privilege.
I'm still half-convinced this is Yegge's version of Andy Kaufman's professional wrestling career, "Gas Wheeltown was an 18 month long sociological experiment conducted by Thiel University. The study is now complete, thank you for your time."
> Jesus. I had to lookup almost every word to make sense of this.
There's a lot to say about this, but I would just start with why are you using NATS if you don't understand it at all, and how did this become AI's fault?
The great attraction of the Ars notoria for John was its promise of quick access to advanced scholarly knowledge. He recounts his continuing wish to study, and how he was lent a copy of a book of necromancy by “a certain cleric”. He copied much of it and wanted more... For John, the Ars notoria was dangerous not in some abstract way, but very directly. He confesses that it struck him at first as beautiful and holy, and seemed to offer miraculous gifts rather than demonic temptations. It became apparent, however, that the book was utterly deceptive, a work of the Devil, a “sick pleasure” that was actually fatally poisonous to the soul, not only to the body.
John records that the Ars notoria was especially attractive because he could not afford all the books needed for his studies, nor the many lectures. He seems to have abandoned his official course, wishing to take advantage of the promise of knowledge of all the sciences, and devoted himself to working with the Ars. ... Further troubling visions ensued, in which arrogant figures demanded worship — something that was clearly against true religion. Gradually it became clear to John that the book hid invocations to demons within what appeared to be beautiful prayers. And yet, like any addict, John found it almost impossible to give up the Ars notoria completely; it was only when the visions became directly life-threatening that the Virgin, with St John the Evangelist as her intermediary, appeared and made it clear that, as John says, “the Ars notoria was deeply evil”.
Beyond the title the analogy is never quite made explicit, but is the subtext that AI "is futile because it cannot deliver what it promises ... its 'signs' are neither understood by humans (like ordinary words and letters) nor sent by God (as sacraments are), and thus are precisely the type of thing that lures humans into contact and compact with demons"? Or that denouncements of AI are coming from the modern equivalent of medieval clergy? Or both?
Not World War 3 though, currently.
https://www.theguardian.com/world/2026/jul/26/ukraine-war-br...
Identifiers included in uploaded data. Information you upload to use the Services may include personal identifiers such as your name and email address (including names and email addresses of repository contributors), as well as other identifiers that may appear in commit metadata, file paths, transcripts, or other uploaded materials.
AI coding-agent transcript-derived content. Per-session narratives, certain session events including file paths and truncated command text, prompt excerpts, steering traces, and dispatch metadata. This data may include local file paths (e.g., directory structures such as /Users/…) and other contextual information that could be associated with an individual. Raw conversation history, full prompts, full agent responses, and full tool outputs do not leave your machine.
Information you upload may include Personal Information relating to other individuals (for example, repository contributors). You represent that you have the necessary rights and permissions to provide such information to us for processing in connection with the Services.
Usage patterns and activity intervals. We may derive and store information about your usage patterns, such as session timing and working-hour intervals, based on timestamps associated with your activity.
Remember that Bobbin "is only a glorified edge index, in the graph theory sense. Additionally it has a record cache, re-filled on demand. All other data that Bobbin serves comes live from PDSes & knots."
You want to first nixery bao proxy spindle with AppRole credentials, then xrpc your punchcard via ATProto. The docs explain all of this clearly, so be sure to read them thoroughly.
Do I get to claim I made the Python TheseusA, or TheseusB, or neither or both?
Just now learning Grigori Rasputin died of malaria.
Sir Thomas More: "Yes! What would you do? Cut a great road through the law to get after the Devil?"
William Roper: "Yes, I’d cut down every law in England to do that!"
Sir Thomas More: "Oh? And when the last law was down, and the Devil turned ’round on you, where would you hide, Roper, the laws all being flat? This country is planted thick with laws, from coast to coast, Man’s laws, not God’s! And if you cut them down, and you’re just the man to do it, do you really think you could stand upright in the winds that would blow then? Yes, I’d give the Devil benefit of law, for my own safety’s sake!"
"quite easy to spot" this style in what sense? To a human? To an LLM? To a deterministic code scanner you wrote by hand yourself, per Codeberg policy?
I'm trying to imagine a world in which this makes sense as a way to organize software development work, but coming up short. It's on a blockchain or something though, so there ya go.
https://github.com/block/buzz/blob/main/docs/assets/screensh...
Spend 3 days a week writing Ruby on Rails and 2 days hand rolling x86 assembly. Every web dev I know has been doing this since long before LLMs. Ensures they can keep having high level Rails thoughts.
Err, do they? For what percent of real world use cases?
The database can scale to handle more traffic by adding replicas. An extreme example of this is OpenAI's use of 50 replicas on a single Primary.
So an extreme example is OpenAI needing 50 replicas, but we're doing five blades ... err, we're doing 768 servers because the need arose "pretty quickly"?
When we needed to store a petabyte of data (one million gigabytes), we'd need many more shards
For who? The United States government? How many end-users are running 1PB Postgres database on DBaaS?