Musings on front-end, privacy, usability, rejecting assumptions and whatever else pops into my head. Working on @ontoapp.
Joined October 2023
- Tweets149
- Following33
- Followers16
- Likes73
Pinned Tweet
An innocent grep here, a little sed there and before you know it you are hooked on bash. Be careful out there kids.
Who is building rehab centers for those afflicted by AI induced psychosis? Could do numbers and sell to PE who would automate the whole thing with AI
This is silly, I don’t think any sane person sees this as a flex. What does maximum potential look like here anyway? 10k PRs a week? What are we even talking about?
We at @coderabbitai shipped 5K PRs last week. Our tiny team is shipping more than a thousand-person dev team. And we are nowhere near hitting our maximum potential. If we industrialize software development correctly, 50K PRs a week are totally achievable, if not more. At this point, it’s machines building machines.
We are looking for pioneers to join our software factory initiative. Must meet one of the following criteria to join the project:
- Tastefully shipped at least 1000 PRs in the last 30 days.
- Played at least 500 hours of strategy or simulation games such as Civilization, Age of Empires, Transport Tycoon, Factorio, etc.
This is a once-in-a-lifetime opportunity to define the UX/DS for building and shipping software at scale. You'll solve a broad spectrum of problems, from designing provably correct software using formal methods to achieving operational excellence with agentic operations. You will also get millions of dollars of token budgets!
Please DM me only if you qualify and want to spend your next few months/years on this!
This makes logical sense and I understand it, but I can't help but think what happens when we remove every 'review' surface that ever existed. 'Well Adam nobody reviews anymore what does it matter?' True! but then where will the decision making layer live? Long lived chat histories? .md files? SOPs? Ontology/spec? ...and to that end where will we go to interact with a history of decisions that dictate an entire business that nobody reviews implementation for? @ontoapp is inevitable.
Why are you still using pull requests?
Is it for code review?
If so, do you really think your or your colleagues catch bugs and mistakes that an Opus 5.5 or Fable or Astra won't find? You can literally spawn all three of them and have them check four times each.
(Also, don't tell me you don't have a "stamp plz" Slack channel or group chat somewhere.)
Is it for knowledge sharing?
Surely there's better ways for that now than the PR, no? We have models that can generate explainer videos, presentations, podcasts, games, quizzes, walkthroughs, ... for any piece of code it can reach.
Is CI the reason?
Well ackshually, CI means continuous integration and its original meaning is the *opposite* of having long-lived branches and PRs.
But, leaving that aside, your agents are hopefully already running the tests before they push something, and hopefully they can run as long as they want and don't block you so they can run the whole test suite and generate "irrefutable proof" and end-to-end demo videos and so on... And so they run the tests and confirm and then you push to a branch and everything runs again and then you merge and then it runs again? That doesn't make sense, does it?
Maybe you work in open source and you accept contributions, in which case I can't speak to the value of pull requests. It's been 2 years since my last OSS contribution. But from what I heard, the time of the PR is coming to an end in OSS too?
I honestly don't get why people still stick to PRs.
With agents you can now move 10x faster, you can finally do real continuous integration, and get more confidence in the changes the agent made.
Don't slow yourself down by having all the changes pile up in a PR.
While some people will get clever with CI and tests, there are alternatives to software factories that make compute prices irrelevant while improving coordination
If the ideal ratio is 1:1 and every team in an org is a single engineer and pm, ownership and understanding of the ontology and it's ability to pass hands when someone leaves becomes the single most important thing.
Creating silos and product teams of 1 is great for enabling efficient ai use, but it feels completely unrealistic to me as a path forward for the majority of orgs. Silos help unblock coordination at one level and defers coordination farther upstream.
Company wide surveillance erhh...I mean "shared data" is not a substitute for a proper ontology. It assumes too many things while simultaneously asking you to give up privacy. Process should be explicitly decided and not quietly mined from your conversations.
Getting business rules into context is great, but it's not foundational enough. Everything should be downstream of process and it needs to legible to every person that comes in to contact with it.
imo visibility is still one of the hardest problems to solve in emerging ai solutions. A full telemetry suite or layered vendor stack is not necessarily the right shape for the majority of builders who will have less and less technical knowledge in a few years.
Overall I think AI tooling needs less layers. At Onto we are excited about making your business not just a layer in your stack but the foundation and source of truth from which everything else flows.
Do we still have preferences on styles since nobody is writing or reading code anymore? Thing is when I want to go read it this tells me absolutely nothing about what this looks like and I'm reminded why I started using tailwind so much... You can reconstruct a lot more mentally when you are working in one file/place and know what your classes are doing.
we have ai slop, have we coined skill slop yet? Skills directories with millions of skills is pretty wild and everyday we get new people "open sourcing" lol their skills? Skills as engagement farming? I dunno man... We are going to have billions to choose from in a couple years
Why does Claude insist on writing it's own terrible SVG approximations instead of grabbing what you referenced quite specifically.
I know the data supports the trend in ASO to list all your competitors on your site with comparison data but ultimately I've found a lot of cool products by companies doing this and the tradeoff of sharing alternatives/competitors for free is probably getting understated?
Would love to see @Apple data that shows users who actually would need or prefer these results using finder search. Looking for my tax ein filing (it used to work doing this) and. instead getting 1. a hospital site I've never heard of, followed by a maps POI followed by a wiki style result is insane. I want to search my computer... I already have the internet for other search types.
Time to bring plan mode back, I've noticed some outputs getting worse despite the PR to the contrary and auto-mode by default is not the answer.