given that all your competitors in this space are open source, hard to make the case against it
if you have any special tricks, probably don't worry too much about getting copied if you really want to win in the "devs who care about performance" market
Early != right. No one gets a co-founder because they want some simple (Claude-doable) work done by them!
You get a co-worker for the expertise, knowledge, and skills that they have that you don't. An AI that has a small/simple subset of that can't replace a co-founder.
Will there be more solo founders now? Of course, there were solo founders before, and the barrier to solo is now lower. But that doesn't change the fundamental fact that it's still (even with AI) relatively rare for a single human to know everything (and how to do everything) that a successful start-up requires.
I worked at Meta and we had AGI in the form of very competent ICs - getting anything done was often super painful because "org priority" this and "buy in" that
what type of control would you like? For me the key is to give my agents more useful OS abstractions and I do this via https://lattices.dev - this gives my projects a nice semantic representation and action space for my OS
I don't do everything. Also fully expect Siri and AppIntents will continue to improve
Funny enough - navigate between apps, targeted control of where I want to land in particular app/window etc.
I’m a dinosaur and still read the code that my ai has written and do a pass with my own eyes of other people PRs. Saying and landing precisely in things like “open this PR in VSCode”. Does it require many actions? No. Would it reduce friction? Hell yea. Doing all that while the ide (example, again) will suddenly decide to rebuild the indexes or run extension updates etc is breaking the flow. Being able to jump into fully loaded and prepared place that is ready to go - it would be very good.
I will admit. First it started with the need to use like 2 ides in everyday work (there was a time at Google when IDEA was defunded and eclipse was sole supported Java ide, but I hated eclipse wholeheartedly and used it only for debugging), and now with ai coding I really feel how my fluency with ides is degrading day after day. I remember myself even 5 years ago (and more so 10-15) - I, somehow, much more rely on mouse or things like “find action/command palette” than I did in past.
https://uselinea.com - A reading app with a harness or a harness with reading app. Can turn any highlight/page/section into audio with your favority TTS engine, including Kokoro if you like
I initially built it to help me read AI papers before discovering AlphaXiv but then I just kept building it for the love of the game
Might have some rough edges but one feature I'm excited about is a more personalized notebook LM style audio generation capability for any readable material
(Another reason to build this was to kick the tires on a framework I've been building to power my own apps - https://hudsonkit.com - helps me ship iOS, macOS and web apps with shared primitives)
the loop is going to be interesting as we produce orders of magnitude low signal code, how do Labs comb through everything to train on the true contributions since last training run?
I wonder if one way to monetize certified "skill" will be to syndicate / license your provably well software engineered / tasteful AI engineered code back to the Labs for a fee
there's one of these just about every other week since early this year
i think it's incredible because however saturated I feel this space has become, highly likely the vast majority of the world has no idea how good models have become, let alone local models
I've added local voice support for all apps/webapps I've built this year
Local TTS is on the cusp of a breakout too, basically a year behind ASR imo in terms of adoption, understanding, size and quality
ACP helps with agent interoperability and helps a lot of IDEs and ADEs integrate all the major harnesses seamlessly
zed https://github.com/zed-industries/zed orca https://github.com/stablyai/orca/tree/main/src/main
I haven't looked at the cli implementation but you can get very far with integrating a harness while not invoking the model directly via API