marvin-hansen
Joined 9 years ago
Joined 9 years ago
However, I would like to point out that Apple isn't totally wrong here because the accessibility API unfortunately is way too broadly scoped, and because of that you literally get access to everything on the computer like you you can screenshot listen and and move the cursor... This is completely ridiculous and the proper engineering solution would actually be to phase out the accessibility API and replace it with something that is narrowly scoped so you can grant specific permissions individually.
However, Apple, being Apple, is obviously not doing anything, and instead says no accessibility permission for anything that isn't demonstrable accessible. Now, there are obviously some exceptions because Apple is not particularly well known for applying its rule consistently and granting big exceptions for itself. However, they do have a valid point on privacy and data protection. And I say that as somebody who ended up distributing my MacOS app outside the App Store because I only got approval for iOS.
That said, I would definitely appreciate if Apple would gradually improve its developer program experience, because compared to its hardware lineup, the developer program is nothing short of abysmal.
https://www.macrumors.com/2026/03/30/apple-pulls-vibe-coding...
GH style import. This is a big one I wish proto had in the first place. The entire idea of a proto registry feels reactive to me when, ideally, you want to pull in a versioned shared file to import that is verified by the compiler long before serve or client verifies the payload schema.
Schema validation and compatibility checks on CI. Again a big one and critical to catch issues early.
Enums done right... No further comment required.
I think with some more attention to details e.g. hammering out the gaps some other comments have identified and more language support e.g. Rust, Go, C# this can actually work out over time.
Here is an idea to contemplate as a side gig with your favorite Ai assistant: A tool to convert proto to Skir. Or at least as much as possible. As someone who had to maintain larger and complex proto files, a lot of proto specific pain points are addressed.
The only concern i have is timing. Ten years ago this would have been a smash hit. These days, we have Thrift and similar meaning the bar is definitely higher. That's not necessarily bad, but one needs to be mindful about differentiation to the existing proto alternatives.
I hope this project gains trajectory and community especially from the frustrated proto folks.
That said, I already love the solution. I was unfortunate enough to integrated stripe and the entire webhook hack really causes more problems that it solves not to mention the ludicrous compliance department at Stripe that should not exist anymore..
I hope you guys get real trajectory.
"What Has a Foundation Model Found? Using Inductive Bias to Probe for World Models"
your thesis of Ai's lack of capacity to abstract or at least extract understanding from noisy data was largely experimentally confirmed. I am uncertain though about the exact mechanics b/c as they used LLM's, its not transparent what happened internally that lead to constant failure to abstract the concept despite ample predictive power. One interesting experiment was the introduction of the Oracle that literally enabled the LLM to solve the task that was previously impossible without the oracle, which means, at least its possible that LLM's can reconstruct known rules. They just can't find new ones.
On a more fundamental level, I am not so sure why these experiments and mathematical proofs still are made since Judea Pearl already established about seven years ago in "Theoretical Impediments to Machine Learning " that all correlation based methods are doomed as they fail to understand anything. his point about causality is well placed, but will not solve the problem either.
The question I have though, if we ignore all existing methods for one moment, then what makes you so sure that AGI is really Mathematically impossible? Suppose some advancement in quantum computing would allow to reconstruct incomplete information, does your assertion still holds true?
https://arxiv.org/abs/2507.06952 https://arxiv.org/abs/1801.04016
Socials: - github.com/marvin-hansen
Interests: AI/ML, Open Source, Programming
---