Sign in

Hillel

@hillelwayne.com
8.2K followers 104 following 5.1K posts

Developer educator at @antithesis.com. Formal methods, software history, chocolatiering. DMs open. *Logic for Programmers* now out! logicforprogrammers.com Newsletter: buttondown.email/hillelwayne

PostsRepliesMedia
Hillel @hillelwayne.com · 10h
I have a new blog post! This is a response to the current formal methods hype and explains what TLA+ is good for and what it's not designed to check. buttondown.com/hillelwayne/...
buttondown.com
What TLA+ can and can't check
Let's chill just a little bit on the "TLA+ will save AI from itself" narrative
0252
Hillel @hillelwayne.com · 28/09/2026
I created an AI model that takes ANY software system and produces a TLA+ model that satisfies ALL its properties. Software engineering is solved forever!
A Python program that just outputs the null TLA+ spec, which automatically satisfies all properties
2312
Hillel @hillelwayne.com · 28/09/2026
I've come to believe that the main use case for property testing is checking invariants and safety properties, so it doesn't overlap as much with unit tests as I originally thought
010
Hillel @hillelwayne.com · 25/09/2026
I thnk it was an argument with brian based on him saying that partition/property testing wasn't useful. I think when I tried to property test the comparison problem I ran into a bunch of issues and wrote on Twitter "man, I really self-owned myself here"
110
Reposted by Hillel
Steve Klabnik @steveklabnik.com · 21/09/2026
Arguing about arguments steveklabnik.com/writing/argu...
steveklabnik.com
Arguing about arguments
Here’s some Rust code: // define a function fn foo (x : i32 , y : i32 ) -> i32 { // body elided } // call it let z = foo ( 5 , 6 ); In programming language jargon, we call x and y parameters and 5 and 6 arguments . Rust does not have many fancy features related to parameters and…
21928
Hillel @hillelwayne.com · 23/09/2026
Inactive
001
Hillel @hillelwayne.com · 23/09/2026
So, uh, apparently twitter is now selling old handles to the highest bidder
481
Hillel @hillelwayne.com · 23/09/2026
Let's gooooooo
040
Hillel @hillelwayne.com · 23/09/2026
This is in under TWO HOURS
190
Hillel @hillelwayne.com · 22/09/2026
Oh hell yeah, thought you already left!
000
Hillel @hillelwayne.com · 22/09/2026
Was thinking the same thing D:
120
Hillel @hillelwayne.com · 22/09/2026
After a move experts are calling "incomprehensible", "ill-advised", and "to New York City", I'm now down to meet tech people in NYC if they want to ask questions and brainstorm and stuff
1230
Hillel @hillelwayne.com · 22/09/2026
I once interviewed ex-"real" now-software engineers for what they thought was better about SE, and EVERY SINGLE ONE said version control
040
Hillel @hillelwayne.com · 22/09/2026
Tomorrow, 9 AM PST, my Systems Distributed talk "Logic for Programmers" goes live! I'll be answering questions and giving background info on the stream, so check it out! @tigerbeetle.com m.youtube.com/watch?v=8Xrv...
070
Hillel @hillelwayne.com · 17/09/2026
Another person to talk to would be @bugarela.com— her Quint language is supposedly very good at bridging the code-spec gap
142
Hillel @hillelwayne.com · 17/09/2026
I should disclaimer that the way we FM folk think about "specifications" doesn't naturally fit into how people intuitively think of "LLM specifications", so I wouldn't be surprised if the main outcome is "inspiration for similar approaches", with the actual techniques developing independently
130
Hillel @hillelwayne.com · 17/09/2026
Another hot topic is studying whether a codebase matches a specification or not, or if one spec is a more detailed version of another. I've not done a lot of this but I know @dominiktornow.bsky.social has a three step process: formal spec to more-detailed formal spec to implementation code
160
Hillel @hillelwayne.com · 17/09/2026
Only half serious, but a lot of the practice of formal specification boils down to "how much can we simplify our model of the system and still have something useful enough for analysis?" Like do we have to specify that a particular batch job runs nightly, or is "it runs sometimes" enough?
150
Hillel @hillelwayne.com · 17/09/2026
*Crashes through skylight* did someone say formal methods?!
1130
Reposted by Hillel
dorian @doriantaylor.com · 17/09/2026
there was a great essay by @hillelwayne.com a while back about how uml never had a standard symbolic representation; the standard only defined the pictures: buttondown.com/hillelwayne/...
buttondown.com
Why UML "Really" Died
There's this post going around the internet called Has UML died without anyone noticing?, In the piece, Ernesto Garbarino says that UML was killed by...
7232
Reposted by Hillel
Jerry Chen @jcsalterego.bsky.social · 17/09/2026
specs get easier once you get the basic idea that code is a homeomorphic endofunctor mapping submanifolds of a Hilbert space
5866
Hillel @hillelwayne.com · 17/09/2026
Has anyone used Ted Chiang's "Catching Crumbs from the Table" as an analogy to AI's progress in math? gwern.net/doc/fiction/...
gwern.net
081
Hillel @hillelwayne.com · 17/09/2026
I feel like Metamorphic testing, checking how changing an input changes a function, would be a pretty good way to evaluate AI tooling. Like if we're writing an AI writing detector, for any random given piece of prose, asking Claude to "clean it up" should *increase* the score
2100
Hillel @hillelwayne.com · 17/09/2026
Someone on LinkedIn has extremely low expectations of my wife, my grandmother, and Donald Knuth
A post saying "You called for Al generated comments to be autobans on
Linkedln. Is that what you want to happen if your spouse posts Al
generated content? How about a grandparent? Donald Knuth?" (I did not call for autobans)
5640
Hillel @hillelwayne.com · 16/09/2026
New Newsletter! Remember that old article about how LLMs liked the word "delve"? Well, the coding harnesses really, REALLY like the word "spine". And "gate". And many others. buttondown.com/hillelwayne/...
buttondown.com
The LLMs yearn for the spines
You can't get them to stop talking about spines!
0150
Hillel @hillelwayne.com · 15/09/2026
Right now the biggest barrier to AI-driven formal methods is that AIs are absolutely dogshit at coming up with good properties. I thought that was just a March 2026 thing but it seems to be a September 2026 thing too
1242
Hillel @hillelwayne.com · 10/09/2026
Are you kidding me
Sorry, you can't be Hillel Wayne because you're already in the author program as Hillel Wayne
0181
Hillel @hillelwayne.com · 08/09/2026
I think you'd be better off with a constraint solver than with TLA+. Check out Minizinc
120
Hillel @hillelwayne.com · 08/09/2026
Logic for Programmers got its first five star rating on Amazon holy shit I know I shouldn't be as elated about this as I am but it's one more sign people are *actually reading the book*
0342
Hillel @hillelwayne.com · 03/09/2026
Why no type theory chapter? A few reasons, but mainly because everything else in the book works off classical logic and I didn't want to spend my page budget on constructive logic. And because much of the "low hanging fruit" applications don't need logic, so a "logic" chapter would get too advanced
0120
Hillel @hillelwayne.com · 03/09/2026
The book discusses some type system techniques and has a short chapter on logic programming, but most of it is on more mundane uses of logic: testing, refactoring, state invariants, case analysis, etc. I'm aiming to get people thinking "oh I can use this right now on the code I already have"
1140
Hillel @hillelwayne.com · 03/09/2026
Whenever I talk about *Logic for Programmers* in public, lots of people assume it's about one of two topics: - Type theory and Curry-Howard - Logic programming and Prolog Those two subjects have so completely dominated the public perception of "math in software engineering"
1251
Hillel @hillelwayne.com · 02/09/2026
This job is the first time I've had unfettered access to Claude code. So, uh, how do people like to customize their Claude rigs? Special skills, hooks, other stuff? I'm not great at using programs in their vanilla state
11120
Hillel @hillelwayne.com · 01/09/2026
*Logic for Programmers* has been out for a month! To celebrate, I'm released the whole second chapter, "A Crash Course in Logic", for free online. That's 7000 words teaching practical math for the working programmer: www.hillelwayne.com/post/predica...
hillelwayne.com
A Crash Course in Predicate Logic
I started writing Logic for Programmers because there weren’t any good resources on logic for, uh, programmers. Now that the book’s out, the new problem is that there aren’t any good free resources on...
0163
Hillel @hillelwayne.com · 01/09/2026
Every picture I see makes me more horrified with the Amazon Direct color printing standards
140
Reposted by Hillel
Predrag Gruevski @predr.ag · 01/09/2026
New programming book arrived in the mail 👀 There's never been a better time to think deeply about the structure behind the code we cause to exist.
Logic for Programmers, a book by Hillel Wayne
1446
Hillel @hillelwayne.com · 25/08/2026
Tired: software engineering isn't dead because Anthropic is hiring lots of software engineers Wired: writing isn't dead because the Claude Code docs are clearly not vibeslopped
4171
Hillel @hillelwayne.com · 19/08/2026
Yep, NixOS. Not hating it so far, but also avoiding doing anything that's actually hard so far
010
Hillel @hillelwayne.com · 19/08/2026
Everybody's favorite part of publishing a book is writing errata
2130
Hillel @hillelwayne.com · 18/08/2026
I think the core engineering teams still are onsite
130
Hillel @hillelwayne.com · 18/08/2026
I do realize the hilarious irony about arguing that "the consumer model is better for developers" in a post where I also admit I'm trying to master NixOS
1100
Hillel @hillelwayne.com · 18/08/2026
This has shades to me of the Lisp Curse: when a language is able to do too much, collaborating with other users of the language becomes a lot harder. Tooling and library authors especially have their work cut out for them. And since tooling and libraries are critical to software adoption, well...
1142
Hillel @hillelwayne.com · 18/08/2026
First true newsletter in two months! This is the Vim vs VSCode article I've wanted to write for a while, about how "Vim mode" in VSCode isn't quite the same as running actual vim, and how ultimately, the "consumer model" is better for most developers. buttondown.com/hillelwayne/...
buttondown.com
Vim wants you to control, VSCode wants you to consume
The two sides of the editor divide, and why I'm on the losing side.
5234
Hillel @hillelwayne.com · 18/08/2026
BTW I work for @antithesis.com now (as of last week)
8990
Hillel @hillelwayne.com · 18/08/2026
I ain't opening this mac
020
Hillel @hillelwayne.com · 18/08/2026
It sounds like a joke but yeah that something you could conceivably do with Minizinc `geost` constraint predicates
010
Hillel @hillelwayne.com · 18/08/2026
Amazon seems to be selling Logic for Programmers for 15% off if you want to go get it for slightly cheaper. I had no idea this was happening and have no control over the sale, but I get the same royalties either way! www.amazon.com/dp/B0HBLP4B26
amazon.com
Logic for Programmers: Wayne, Hillel: 9798995302902: Amazon.com: Books
Logic for Programmers [Wayne, Hillel] on Amazon.com. *FREE* shipping on qualifying offers. Logic for Programmers
1132
Hillel @hillelwayne.com · 17/08/2026
I got a macos too from work but I'm not gonna master nixos if I don't go all in
100
Hillel @hillelwayne.com · 17/08/2026
work laptop
000
Hillel @hillelwayne.com · 17/08/2026
nix-shell -p --run is so cool
130