John Cheng

Programming since 2000

[Andrej Karpathy @karpathy]

This is the the quote I've been citing a lot recently.

[Gergely Orosz @GergelyOrosz]

Congrats on having fun and building a vibe coded WYSIWYG editor

Patiently waiting for when they’ll realize things like permissions, tagging, backups+disaster recovery, search, exports, tables, mobile app, integrations w Slack+Linear+others need to be built…

🍿

[Andrej Karpathy @karpathy]

Bought a new Mac mini to properly tinker with claws over the weekend. The apple store person told me they are selling like hotcakes and everyone is confused :)

I'm definitely a bit sus'd to run OpenClaw specifically - giving my private data/keys to 400K lines of vibe coded monster that is being actively attacked at scale is not very appealing at all. Already seeing reports of exposed instances, RCE vulnerabilities, supply chain poisoning, malicious or compromised skills in the registry, it feels like a complete wild west and a security nightmare. But I do love the concept and I think that just like LLM agents were a new layer on top of LLMs, Claws are now a new layer on top of LLM agents, taking the orchestration, scheduling, context, tool calls and a kind of persistence to a next level.

Looking around, and given that the high level idea is clear, there are a lot of smaller Claws starting to pop out. For example, on a quick skim NanoClaw looks really interesting in that the core engine is ~4000 lines of code (fits into both my head and that of AI agents, so it feels manageable, auditable, flexible, etc.) and runs everything in containers by default. I also love their approach to configurability - it's not done via config files it's done via skills! For example, /add-telegram instructs your AI agent how to modify the actual code to integrate Telegram. I haven't come across this yet and it slightly blew my mind earlier today as a new, AI-enabled approach to preventing config mess and if-then-else monsters. Basically - the implied new meta is to write the most maximally forkable repo and then have skills that fork it into any desired more exotic configuration. Very cool.

Anyway there are many others - e.g. nanobot, zeroclaw, ironclaw, picoclaw (lol @ prefixes). There are also cloud-hosted alternatives but tbh I don't love these because it feels much harder to tinker with. In particular, local setup allows easy connection to home automation gadgets on the local network. And I don't know, there is something aesthetically pleasing about there being a physical device 'possessed' by a little ghost of a personal digital house elf.

Not 100% sure what my setup ends up looking like just yet but Claws are an awesome, exciting new layer of the AI stack.

[Matt McDermott @mattmfm]

A huge political earthquake in Texas tonight as Democrats flipped a State Senate seat from red to blue in a district Trump won by 17 points.

Trump personally waded in — endorsing the Republican and personally urging base turnout — and was dealt a massive loss.

[Daniel Eth (yes, Eth is my actual last name) @daniel_271828]

Feels like the dam has broken on people in the tech community airing grievances with Andreessen. Honestly makes me feel better about the direction of the tech community writ large

[Gergely Orosz @GergelyOrosz]

The promise of AI chat assistants: they solve 90% of the problems users have (by looking up the docs and telling them)

My reality: need to spend 10 minutes trying to get to a human, to solve an issue I need customer support to look into

Around minute 8 I sign up to a competitor

[Vicent Martí @vmg]

This is one of the things I’ve been doing at Cursor! We’ve migrated our largest database workload, the `codebase` cluster, from AWS Aurora Limitless to PlanetScale. The cluster contains syncing metadata for all the user codebases that we index.

Why did we migrate from Aurora Limitless? Basically, it’s a very poor database offering. AWS could tell that the Postgres market is very hot right now and they wanted to be first-to-market with a “sharding solution” for infinite scalability. But the product they’ve launched is veeeery far from production ready. From availability, to performance to pricing… It was just a constant stream of issues. And you know with AWS, when the system works it works quite well, but when it breaks, it’s very hard to get good support, and that made operating the cluster a nightmare. Limitless just turned out to actually have a lot of limits in practice.

Why did we choose PlanetScale? I did try to stay out of the conversation, because it felt kinda lame to start pitching my previous startup right after joining a new one, and obviously I’m biased because I’ve worked really hard these past 5 years on performance for the PlanetScale platform. Overall, I think all the infra folks already had a lot of confidence on the capacity for scale Vitess has, so it boiled down to whether it actually made sense to port our Postgres workload to MySQL.

Ironically, what made the choice easier is that Aurora Limitless **doesn’t support replication streams** (seriously!! it’s in the roadmap but right now you’re just locked-in if you move your data in!!!), and since we had to make major changes to perform the migration at the application level, it wasn’t a significant effort to move from Postgres to MySQL while we were at it.

So yes, the migration was not easy, but the end result was 100% worth it. Despite MySQL’s many shortcomings, the performance of Vitess + MySQL at scale on PlanetScale Metal is just unreal, miles ahead of anything you can do with a Postgres-compatible database right now. Can’t wait to see what the PS team does with their sharding solution for Postgres — it feels like a massive gap on the database landscape right now.

[Armand Domalewski @ArmandDoma]

I do not want to live in a country where we resolve political disputes with bullets and blood

[Mason @webdevMason]

I've met exactly two kinds of parents I truly admire on this axis, and one of them never, ever lies to their kids

The other tells their kids they're made of lego and if you don't come here right now I'm taking your arms off

[Gergely Orosz @GergelyOrosz]

It’s cringe to see successful companies that have nothing to do with AI rebrand themselves as “an AI platform” because they think it will somehow help them.

Workday is a HR platform.

Sure, they added AI features like everyone else but was never “built with AI at the core”

[John Cheng @jlicheng]

Why do you hate Jira?

[Gergely Orosz @GergelyOrosz]

How is it that Atlassian builds tools that:

- SO MANY people use

- SO MANY people dislike (JIRA, Confluence. Previously also HipChat)

They are defying the conventional wisdom for business success that starts with "build something people will love"

[John Cheng @jlicheng]

Why is this even a thing?

[Gergely Orosz @GergelyOrosz]

Annoying how global companies don't think about the use case of "what if a customer wants to use the same email for their account in two different countries."

If you buy an LG smart appliance, you need to use a different email in every country.

Amazon solved this: LG did not

[Andrej Karpathy @karpathy]

I am (slowly) re-reading the Tolkien legendarium (of which Lord of the Rings is a small part). The whole body of work is so incredible and there's nothing else like it... it dilutes other worlds of fiction. Wait - your story doesn't have a comprehensive history/mythology spanning multiple ages all the way back to a creation myth as detailed in separate volumes? You didn't first invent new languages and dialects for your characters? You didn't pack it with powerful themes and stories written it in a beautiful, archaic style and compose poems and songs alongside? It didn't take you multiple decades of iteration? And what of all the uncharted territory still remaining? Is Tom Bombadil one of the Ainur. Where are the Entwives. What happened to the two unaccounted Istari. Can we hear more about what it was like in Cuiviénen when the elves first awoke? Or to see the light of the two trees of Valinor. Or of the splendor of the caves of Aglarond.

What's most on my mind though - the Tolkien legendarium is imo a concrete example of a height of culture. Does AI, today or soon, make it easier to reach this high via empowerment in both writing and ideation? Or harder, when quick wins are tempting and ~free, and an independent ability to create is stifled. If such a body of work is made again but now with heavy AI assistance, does it inspire the same wonder? What if thousands of them come out on demand with just a prompt? Why do you feel cheated when you learn that something your read was AI generated? Is it transient or a function of capability? Is it slop? What is slop? Or is wonder inseparable from its own creation myth of a lifelong obsession of a mind like your own? So many questions.

[John Cheng @jlicheng]

Thank goodness that Google Docs has a Copy as Markdown option. As it allows me to paste my doc into a real LLM (the chat version of Gemini, even), instead of the crappy version embedded inside GDocs.

[Bill Maher @billmaher]

Among Democratic primary voters, 16 % would support Pete Buttigieg. But among black voters, zero.

[Gergely Orosz @GergelyOrosz]

Had a few typos slip through in my newsletter (sorry!) and could not understand why. This is after 2 rounds of manual editing + 1 round of final spellcheck with Grammarly.

Well, Grammarly cannot catch obvious typos.

Claude, meanwhile, can.

Switching Grammarly for Claude. Wow.

[Gergely Orosz @GergelyOrosz]

The evolution of "vibe coding production software as a non-dev" the last few months

Success stories do exist... ones I see all come from people who are/were software engineers

Learning to code / build software not a waste after all?

[John Cheng @jlicheng]

"Ready" is doing a lot of the work there

[Gergely Orosz @GergelyOrosz]

An example of when something sounds too good to be true it probably is:

“A platform where you can generate AI-powered, production ready apps.”

Lots of vibe coding platforms make this claim: but none can actually deliver on it. It’s because of a mix of things (cont’d):

[Jay Bookman @jaysbookman]

ICE is deporting a permanent legal US resident from Korea, a scientist researching a Lyme disease vaccine who has lived here since age 5, citing a minor pot possession charge 14 years ago for which he did community service.
My fellow Americans, this is absolute madness.

[Spencer Hakimian @SpencerHakimian]

Well would you look at that.

2 *perfectly* timed trades shorting the Dollar and then longing the Dollar *right before* Trump leaked that he was firing Powell, and then TACO'd 30 minutes later.

10:54 AM (first red circle), heavy shorting pressure on the USD/EUR begins out of nowhere.

It then kicks into overdrive at 11:13 AM (second red circle).

And then, at 11:16 AM (green arrow), alerts flash on screens worldwide that Trump is planning to fire Powell as soon as today.

And the Dollar goes into freefall.

But then, while markets are trying to digest this bombshell, the Dollar starts *rallying* rapidly versus the Euro at 11:54 AM (blue circle).

And then, at 11:56 AM, Trump announces that he doesn't intend on firing Powell (purple circle), but by then, the dollar rally is already over with.

I am sure this is just another coincidence.