frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: MacMind – A transformer neural network in HyperCard on a 1989 Macintosh

https://github.com/SeanFDZ/macmind
51•hammer32•3h ago
I trained a transformer in HyperCard. 1,216 parameters. 1989 Macintosh. And yes, it took a while.

MacMind is a complete transformer neural network, embeddings, positional encoding, self-attention, backpropagation, and gradient descent, implemented entirely in HyperTalk, the scripting language Apple shipped with HyperCard in 1987. Every line of code is readable inside HyperCard's script editor. Option-click any button and read the actual math.

The task: learn the bit-reversal permutation, the opening step of the Fast Fourier Transform. The model has no formula to follow. It discovers the positional pattern purely through attention and repeated trial and error. By training step 193, it was oscillating between 50%, 75%, and 100% accuracy on successive steps, settling into convergence like a ball rolling into a bowl.

The whole "intelligence" is 1,216 numbers stored in hidden fields in a HyperCard stack. Save the file, quit, reopen: the trained model is still there, still correct. It runs on anything from System 7 through Mac OS 9.

As a former physics student, and the FFT is an old friend, it sits at the heart of signal processing, quantum mechanics, and wave analysis. I built this because we're at a moment where AI affects all of us but most of us don't understand what it actually does. Backpropagation and attention are math, not magic. And math doesn't care whether it's running on a TPU cluster or a 68030 from 1989.

The repo has a pre-trained stack (step 1,000), a blank stack you can train yourself, and a Python/NumPy reference implementation that validates the math.

Comments

gcanyon•1h ago
It's strange to think how modern concepts are only modern because no one thought of them back then. This feels (to me) like the germ theory being transferred back to the ancient greeks.
hammer32•1h ago
Right? Backprop was published in 1986, a year before HyperCard shipped. Attention is newer, but a small model like this was buildable.
anthk•48m ago
Lisp is from 1960's and with s9 you can do even calculus with ease, in an interpreter small enough to fit in two floppies.

On the Greeks, Archimede almost did 'Calculus 0.9'.

kdhaskjdhadjk•18m ago
I think it's incredible to see the potential that is still locked up in old hardware. For example the 8088 MPH demo. Amazing what he was able to do with an 8088 and CGA. All this time the hardware had that potential, but it took decades to figure out how to unlock it, long after the hardware was considered obsolete. Imagine the sort of things that might be done later down the road with hardware of 0-20 years ago if somebody really dug into it to that level.
DetroitThrow•1h ago
This is very cool. Any more demos of inference output?
hammer32•5m ago
Thanks! The quickest way to try it is the HyperCard Simulator link someone just posted in this thread: https://hcsimulator.com/imports/MacMind---Trained-69E0132C — go to the Inference card, click New Random to fill in 8 digits, then click Permute. The model predicts the bit-reversed permutation of all 8 positions. The pre-trained stack gets all inputs correct.
hyperhello•1h ago
Hello, if there are no XCMDs it should work adequately in HyperCard Simulator. I am only on my phone but I took a minute to import it.

https://hcsimulator.com/imports/MacMind---Trained-69E0132C

hammer32•7m ago
I had no idea your simulator existed. No XCMDs, correct; everything is pure HyperTalk. I just ran a few training steps and they complete in a second or two. Thank you for importing it!
hyperhello•2m ago
I gotta ask. Your scripts have comments like -- handlers_math.hypertalk.txt at the top. Are you using some kind of build process for a stack?

Claude Opus 4.7

https://www.anthropic.com/news/claude-opus-4-7
633•meetpateltech•2h ago•497 comments

Cloudflare Email Service

https://blog.cloudflare.com/email-for-agents/
236•jilles•3h ago•102 comments

Mozilla Thunderbolt

https://www.thunderbolt.io/
207•dabinat•3h ago•178 comments

We gave an AI a 3 year retail lease and asked it to make a profit

https://andonlabs.com/blog/andon-market-launch
48•lukaspetersson•1h ago•60 comments

Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All

https://qwen.ai/blog?id=qwen3.6-35b-a3b
443•cmitsakis•3h ago•227 comments

IPv6 traffic crosses the 50% mark

https://www.google.com/intl/en/ipv6/statistics.html?yzh=28197
636•Aaronmacaron•1d ago•432 comments

Launch HN: Kampala (YC W26) – Reverse-Engineer Apps into APIs

https://www.zatanna.ai/kampala
22•alexblackwell_•1h ago•19 comments

Put your SSH keys in your TPM chip

https://raymii.org/s/tutorials/Put_your_SSH_keys_in_your_TPM_chip.html
25•type0•4d ago•10 comments

Cloudflare's AI Platform: an inference layer designed for agents

https://blog.cloudflare.com/ai-platform/
96•nikitoci•3h ago•22 comments

The future of everything is lies, I guess: Where do we go from here?

https://aphyr.com/posts/420-the-future-of-everything-is-lies-i-guess-where-do-we-go-from-here
227•aphyr•3h ago•217 comments

Show HN: MacMind – A transformer neural network in HyperCard on a 1989 Macintosh

https://github.com/SeanFDZ/macmind
51•hammer32•3h ago•10 comments

Darkbloom – Private inference on idle Macs

https://darkbloom.dev
406•twapi•12h ago•197 comments

AI cybersecurity is not proof of work

https://antirez.com/news/163
133•surprisetalk•5h ago•57 comments

The paper computer

https://jsomers.net/blog/the-paper-computer
216•jsomers•3d ago•65 comments

Codex Hacked a Samsung TV

https://blog.calif.io/p/codex-hacked-a-samsung-tv
143•campuscodi•6h ago•80 comments

Laravel raised money and now injects ads directly into your agent

https://techstackups.com/articles/laravel-raised-money-and-now-injects-ads-directly-into-your-agent/
111•mooreds•1h ago•64 comments

FSF trying to contact Google about spammer sending 10k+ mails from Gmail account

https://daedal.io/@thomzane/116410863009847575
293•pabs3•13h ago•178 comments

Claude Opus 4.7 Model Card

https://anthropic.com/claude-opus-4-7-system-card
88•adocomplete•2h ago•39 comments

Six Characters

https://ajitem.com/blog/iron-core-part-2-six-characters/
20•Airplanepasta•3d ago•0 comments

Modern Microprocessors – A 90-Minute Guide

https://www.lighterra.com/papers/modernmicroprocessors/
124•Flex247A•4d ago•15 comments

€54k spike in 13h from unrestricted Firebase browser key accessing Gemini APIs

https://discuss.ai.google.dev/t/unexpected-54k-billing-spike-in-13-hours-firebase-browser-key-wit...
330•zanbezi•4h ago•236 comments

ChatGPT for Excel

https://chatgpt.com/apps/spreadsheets/
286•armcat•19h ago•174 comments

Ancient DNA reveals pervasive directional selection across West Eurasia [pdf]

https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/2026_Akbari_Nature_s...
48•Metacelsus•5h ago•35 comments

PHP 8.6 Closure Optimizations

https://wiki.php.net/rfc/closure-optimizations
53•moebrowne•2d ago•8 comments

RamAIn (YC W26) Is Hiring

https://www.ycombinator.com/companies/ramain/jobs/bwtwd9W-founding-gtm-operations-lead
1•svee•9h ago

Cybersecurity looks like proof of work now

https://www.dbreunig.com/2026/04/14/cybersecurity-is-proof-of-work-now.html
513•dbreunig•1d ago•193 comments

Artifacts: Versioned storage that speaks Git

https://blog.cloudflare.com/artifacts-git-for-agents-beta/
26•jgrahamc•3h ago•0 comments

RedSun: System user access on Win 11/10 and Server with the April 2026 Update

https://github.com/Nightmare-Eclipse/RedSun
143•airhangerf15•12h ago•37 comments

North American English Dialects

https://aschmann.net/AmEng/
111•skogstokig•13h ago•64 comments

Too much discussion of the XOR swap trick

https://heather.cafe/posts/too_much_xor_swap_trick/
122•CJefferson•3d ago•93 comments