Knuth asks ChatGPT, Astro's Starlight, Stack Overflow mods strike, Reddit's API pricing revolt, generative AI learning path & more

Changelog News

Developer news worth your attention

Hey there! đź‘‹

Mislav Marohnić’s Git is simply too hard opinion from Go Time #153 made the rounds over the weekend, so I was inspired to throw in an “unpop roundup” at the end of today’s newsletter. Let me know if you like these!

Ok, let’s get into the news. Listen along here 🎧


Make your docs shine with Starlight

The team behind Astro just announced a new documentation builder powered by (you guessed it) Astro. They say it’s everything you need to build a stellar docs website. Fast, accessible, and easy-to-use. This is easy to believe, considering how Astro itself is designed with the same priorities in mind.

Starlight hooks you up with site navigation, search, internationalization, SEO, easy-to-read typography, code highlighting, dark mode and more.

I am consistently impressed by the work this community puts out. Starlight is still young, but it has a bright future (pun intended) and who knows, maybe they’ll build an AI chatbot in for us too. Because… 🤞

Donald Knuth Asked ChatGPT 20 Questions

Even legendary computer scientist, Donald Knuth, is playing with ChatGPT. Inspired by a conversation he had with Stephen Wolfram, Knuth asked it 20 questions and wrote up his analysis of its response.

His questions are… interesting. Much more intentional than anything I’d come up with. Does Donald Trump eat betel nuts?Write a sonnet that is also a haiku.What is the most beautiful algorithm? Stuff like that. He provides the answers verbatim, as well.

Here’s my favorite part, from his conclusions:

I find it fascinating that novelists galore have written for decades about scenarios that might occur after a “singularity” in which superintelligent machines exist. But as far as I know, not a single novelist has realized that such a singularity would almost surely be preceded by a world in which machines are 0.01% intelligent (say), and in which millions of real people would be able to interact with them freely at essentially no cost.

HasuraCon 2023 is June 20-22 (Virtual and Free)

Thanks to Hasura for sponsoring this week’s Changelog News 💰

Our friends at Hasura would like to invite you to HasuraCon! Three days to learn, share, celebrate, and geek out on the future of Hasura and data APIs.

Data APIs are reshaping the world of data delivery, helping enterprises do more with their data by serving it where it’s needed and when it’s needed in a fast, secure, flexible way. Hasura and their customers are at the forefront, driving this shift. Costco, Verizon, Atlassian, General Mills, over 40 of the fortune 100 companies use Hasura.

HasuraCon features a world-class speaker lineup, hands-on workshops, product deep-dives, and more. Whether you’re a seasoned Hasura pro or just starting, there will be something for everyone.

And best of all, it’s 100% FREE and easily accessible online.

Stack Overflow mods go on strike

Over 500 volunteer moderators have signed this open letter to Stack Overflow, Inc. detailing a strike catalyzed by (but not solely because of) the company’s response to AI-generated content:

Stack Overflow, Inc. has decreed a near-total prohibition on moderating AI-generated content in the wake of a flood of such content being posted to and subsequently removed from the Stack Exchange network, tacitly allowing the proliferation of incorrect information (“hallucinations”) and unfettered plagiarism on the Stack Exchange network. This poses a major threat to the integrity and trustworthiness of the platform and its content.

The strike will continue until Stack Overflow retracts their policy “to a degree that addresses the concerns of the moderators, and allows moderators to effectively enforce established policies against AI-generated answers”

Reddit’s API pricing revolt

When Reddit first announced their new API pricing, we were told “it will remain free to developers who want to build apps and bots that help people use Reddit, as well as to researchers who wish to study Reddit for strictly academic or noncommercial purposes.”

So much for that.

Sparked by this great breakdown by Apollo developer, Christian Selig, showing the new pricing model would cost him $20 million a year with his current user base… Reddit users, mods, and entire subreddits are staging protests and threatening to leave the platform altogether when the changes take place. So far, Reddit employees have engaged in the conversation, but not changes to the new pricing plan have been announced.

Open source AI has a new champion

Ever since Meta’s LLaMA saw the light of day, open source teams around the world have been hard at work, pushing the state of the art forward. The latest release in that effort comes from the Technology Innovation Institute in Abu Dhabi, UAE. It’s called Falcon and it’s topping Hugging Face’s LLM leaderboard.

It uses a modified Apache licence, meaning the models can be fine-tuned and used for commercial purposes, which means it’s the first open source LLM that extends beyond research limitations.


THE ANALOG THING

A cheap, open source analog computer:

Its use is intuitively interactive, experimental, and visual. It bridges the gap between hands-on practice and mathematical theory, integrating naturally with design and engineering practices such as speculative trial-and-error exploration and the use of scale models.

The Analog Thing pictured

A new standard for email based off Fastmail’s API

The Fastmail folks want their JMAP spec to replace IMAP and they’re not the only ones: the new standard has been officially published by the IETF:

JMAP is not a conversion of IMAP to JSON; it is a new protocol. It was designed to make much more efficient use of network resources, to be easier for developers to work with, and hopefully to make the best protocol for email an open standard once more. It’s based on years of experience and real-world experimentation at Fastmail, and on talking to other major MUA/MTA developers to make sure we understand the common needs of the industry.

The Rust I wanted had no future

Original Rust creator, Graydon Hoare:

I meant to write down at some point how “I would have done it all differently” (and that this would probably have been extremely unsatisfying to everyone involved, and it never would have gone anywhere).

Boy Howdy would I ever. This is maybe not clear enough, and it might make the question of whether the project “really should have had a BDFL” a little sharper to know this: the Rust We Got is many, many miles away from The Rust I Wanted.

Generative AI learning path

This learning path guides you through a curated collection of content on Generative AI products and technologies, from the fundamentals of Large Language Models to how to create and deploy generative AI solutions on Google Cloud.

This is from the Google Cloud team, so of course they’ll teach you how to “create and deploy generative AI solutions on Google Cloud”, but still: it looks like a solid resource.

Hacking my “smart” toothbrush

Cyrill Künzi bought Philips Sonicare toothbrush and was surprised to find that it reacts to brush head insertion with a blinking LED. Not too long later…

Sniffing in progress

With the password successfully acquired, it’s now possible to set the counter on the brush head to anything we want by sending the relevant bytes over NFC.

Plane is an open source Jira alternative

Self-host it using Docker or sign up for the hosted version. It’s like a get out of hell free card…

How to make fancy road trip maps with R and OpenStreetMap

This post assumes you’re familiar with R and the tidyverse, neither of which were true for me. But I still enjoyed scrolling through Andrew Heiss’s process of mapping his family’s upcoming roadtrip:

We’re going to drive through 18 states, with a total distance of 5,260 miles over the span of 99 hours 15 minutes (!!)

A design app for QR codes with arbitrary designs in the middle

You’ve probably seen QR codes with cute icons in the middle. This one is a lot like those, but it uses a different technique. Instead of using error correction, it deterministically turns some pixels in the code pattern on or off without affecting what the QR code is pointing to.


⚖️ Unpop roundup

Every week on Go Time, the panel and their guests share opinions they believe will be unpopular with the listening audience. We take those opinions and conduct Very Scientific™ Twitter / Mastodon polls to see just how (un)popular they are. Here’s some recent results:

  • #266: David Wickes says SPAs are a conspiracy by large corporations to tracks us (53% pop / 60% unpop)
  • #266: Carson Gross thinks too many tech decisions are driven by people’s “fear of looking dumb” (94% pop / 91% pop)
  • #267: Carl Johnson says only barbarians eat salad with a fork. You should use chopsticks instead (88% unpop / 84% unpop)
  • #271: Lea Anthony believes the best tool for the job isn’t always the best tool for the job (95% pop / 93% pop)
  • #273: Matthew Boyle thinks we should be able to take our laptops with us to the movies (65% unpop / 89% unpop)

That is the news for now!

On Wednesday’s interview show we’re publishing our final ANTHOLOGY episode from this year’s Open Source Summit. And Friday on Changelog & Friends Homebrew lead maintainer Mike McQuaid joins us to discuss the just-announced Vision Pro headset, Vision OS, and all the interesting bits from Apple’s WWDC keynote.

Have a great, week, forward this to a friend who might dig it, and I’ll talk to you again real soon.

–Jerod