dgunay

Definitely. I read a few articles top to bottom that I normally wouldn't have even bothered to look up on real Wikipedia, learned some cool local lore.

I use Muse. So far the biggest value to me is 1) being a more "squishy" reminder service than a typical task tracker, because it can at its own discretion raise stuff to my attention and 2) attempting things that I normally wouldn't have the energy to do on my own. It doesn't always succeed, but a 50/50 chance of it even getting done is still an improvement in my book. Things like 311-ing small problems I see on walks, grabbing free games from EGS, etc.

Maybe the $25B isn't guaranteed and doesn't seriously harm them on its own, but if the legal landscape shifts enough to make fines of this scale more common than the usual dozens-to-hundreds of millions, then it starts to impact their bottom line.

They're very different games. I'd play 64 first simply because it's a landmark in gaming history, but Paper Mario is one of my favorite games ever.

It may not, by some philosophical definition, but most consumers (and I'd say most programmers too) place little importance on code as a human artistic expression. In the case of games the art assets are also a much more clear and direct signal that there was AI involved in the creation of the game. A game with handmade art but vibe coded game logic might not be as obvious.

Doesn't this have potential applications in disaster relief?

I feel like a curmudgeon because like yes the demo was really cool, but more from the standpoint that they were willing to do a stunt like that in an industry that is incredibly boring (at least in the US). The engineering behind making a relatively heavy object fly around with thrusters is very well understood at this point, I wouldn't call it "incredible" to rig RCS thrusters on a brick and make it do a barrel roll.

EDIT: I want to make it clear that 1) I'm not talking about the driving upside down thing, and 2) just because I am not sensationalizing the kickflip trick as some incredibly complex endeavor doesn't mean I didn't think it was cool and I'm some kind of hollow, depressed sad sack. The kick flip made me laugh my ass off, but be real. His companies make rockets that can belly flop and land themselves, stream hi res video of reentry using LEO satellite constellations, catch huge boosters in towers, and this is what's blowing your mind?

I briefly shadowed a doctor when I was interested in potentially going to medical school. I sat down while he met with a patient. He left the room pretty quickly. From the conversation I figured he'd be right back and just went to get something or check something. He comes back in 5 minutes later and I find out that he'd gotten up and seen 3-4 other patients before realizing I was still in the first room.

Some parts of it are not super orthogonal. Intervals aren't visually consistent, for example (E -> F vs C -> C#).

Having to read notes that are up in the ledger lines outside the clef kind of sucks. You just have fewer visual landmarks to hang off of.

This is partly a consequence of the physical instruments we have and their natural ranges, but clefs as a general concept are annoying to deal with when you have to juggle them constantly.

IMO it doesn't do a good job of communicating dynamics and expression. The notation used is vague (accelerando... okay how much faster, over how many measures, etc).

Maybe I'm holding it wrong, but recently I transcribed some of my friend's songs and managing repeats that don't neatly fit into a rigid song structure is difficult without writing very "WET" music that takes up a lot of space.

In a more meta sense, most sheet music is garbage. There is no inherent pit of success when it comes to creating sheet music, and writing a very good, readable score is something only well-trained musicians and composers can do. Look at free sheet music or read notation written by amateurs, you'll see missing key signatures, misspelled chords, etc. Sheet music doesn't make any concessions for people who understand music as sound first, you have to adhere to tradition and theory to produce good sheet music. Lots of musicians with decent ears cannot do this.

IDK, this is me trying very hard to come up with reasons to hate on sheet music.

We still do code review but it does bottleneck how much we can ship.

I'm looking at what we can do to alleviate this. Here's what I think the near future may look like, as far as getting more value per time spent out of human code review. I'm coming at this from the POV of an org that still mandates human code review and I think a lot of this advice doesn't hold if you are all-in on the black box software factory approach.

If you're reviewing code you're responsible for, your job now is to provide stewardship of its long term technical direction even if you are not personally reading and writing every line of code. This means preserving the coherence of designs past and present, and ensuring that they are not being worked around in service of a short-term goal.

Stop reading every line of every file to look for minor gotchas. Focus on core business logic and whether invariants are modeled well & to the fullest extent possible by the type system. Agents are better than most humans now at spotting insane gotchas. At least one frontier model should review most PRs.

In the before times, you could argue that TODO comments and small nits on PRs should be addressed, because most likely they'd never get attention again once the PR merges. Now, small issues should instead be automatically recorded and followed up on quickly by another agent. Minor code quality issues should become rules, and old code should be constantly revisited by agents to bring it up to par with new code.

Provide the transcripts of the session with the PR. Look for evidence of thoughtful direction & solid premise behind the conversation that led to the PR being made. In addition to providing more mentorship opportunities to the author, this also discourages meat proxying/low-value-added uses of AI. Asking someone to rewrite or completely discard a PR that's just a bad idea should no longer be the ego bruising event it once was (at least with respect to the cost sunk to make the PR).

Require a higher bar for proving that code works and stays working. The testing pyramid IMO should now be more of a testing hourglass. Or an upside-down pyramid, even. Automated E2E tests should be abundant. Every feature should receive adequate manual testing by an agent, with proof of work.

In the vein of "stays working," look for good observability and guardrails around a feature. If you're doing a good job of leveraging agents to keep code up to par, this means adding (and cycling out) feature flags, ensuring that there's adequate logging, having a cohesive monitoring solution with good coverage, and a plan for what to do if there's an issue in prod.

Gemini Flash is very dumb and gimped. Recent experience:

1. I ask it to list out all my old Keep notes that are just throwaway task lists, reminders, etc. to propose deleting them.

2. It comes up with like 4 of them (I know for a fact there are dozens, if not hundreds). I ask it for an exhaustive list.

3. I see its thinking trace briefly show "Refusing to produce an exhaustive list..." and then it says "Yes, that is all of them."

Not sure whether it's lack of intelligence, bad tools, or both, but that was extremely frustrating to see.

Piracy is a service problem! If they release a good PC port, I will buy it, as I have done for nearly every other Rockstar game. Hell, if they were to even just automate the port using a similar tool, I wouldn't really mind if the end result were indistinguishable from a decent PC port. I just really don't want to buy a PS5 for a couple exclusives.

Don't worry, I'm sure they're going to aggressively tackle the cultural and economic problems contributing to their fertility crisis, right? Right?

Well, we do at least have to eat enough food and sleep ~1/3rd of our lives. That's kinda maintenance

The job market should be just that, a market, between buyers and sellers, not people who pretend to be buying in order to look like they've got money.

I know this analogy breaks down a little bit since this behavior does actually happen, sort of, in real life p2p markets but the point is, it sucks and if you've ever sold anything on fb marketplace you know how obnoxious it is dealing with people who demand a lot of you but do not intend to actually commit to buying what you're selling.

Do most people understand why cards have a chip in them now?

There is. Doesn't work on every browser but it's really nice and I use it routinely.

That's nice of you, but I'd urge anyone with influence at Google to consider 1) many many many more people who use Google services probably don't know or run in the same circles as Googlers who can bail them out of such a scenario, and 2) these people, being much less likely to understand 2FA, are also more likely to accidentally lock themselves out. An separate class of people who know how a system works or have connections to its operators become nose blind to its terrible UX because they have access to workarounds.

Do you mean you cut the excess margins, or cut out specific sections, etc?

Haven't touched it in a few years but Bevy was a really nice breath of fresh air when I dipped my toes back into hobby game dev. I wasn't used to being able to create a game almost purely as code without having to touch a WYSIWYG editor.

For things of global relevance, I keep a git repo in ~/.skills. It has a script that syncs skills across agent harnesses when run. If I think of an interesting skill that seems like it'd be broadly useful, I put it in this repo. If I want to use someone else's skill, I read it with my eyeballs, decide whether I like it, and then copy and paste it into the repo.

This is a big reason why I feel like even though LLMs are _effectively_ AGI in some regard, they also are a hack around what most people figured AGI would look like before the advent of LLMs. Humans can do metacognition, output multimodally at the same time (verbal _and_ physical intelligence go together to produce an expressive face while one talks), have a good sense for what they do and don't know, continuously take in and respond to the world around them in a (mostly) uninterrupted fashion without "turns", learn knew knowledge and retain it for their whole lives, etc. When you reduce a human to a text generator, yes obviously SOTA LLMs perform way better, but rather than invent something that can operate as an always-running "being", we've grafted a harness around an intelligence that is bound purely to speak only when spoken to. Maybe organic intelligence is already that, playing out at a super high refresh rate, but I don't know.

Luna max is suitable for like 90% of the kinds of code changes I want to make. I only find myself actually reaching for a Sol or Fable tier model if the problem is very complex. If you're willing to build the guardrails and do some extra planning, Luna is very capable.

Everyone has their own distinct mix of jargon and uses some words more or less than others, but now it's as if you have 1000 coworkers who are all the same person, and talk the exact same way, and it just so happens that their writing is grating to read. It also doesn't stop when you log off for the day. The same guy follows you home in ad copy, message board comments, spam, and in some cases even the correspondence of your loved ones.

Oh definitely, there's great value in it for those who already own Macbooks for one reason or another. I just personally have found myself shying away from buying a Macbook multiple times over the years because the Asahi Linux experience was really really rough at the time & probably still is for the generations I would be interested in buying.

I have immense respect for the incredible work that the Asahi team does, but at the rate that Intel and AMD have been catching up on power efficiency I wonder whether anyone will really _have_ to run Linux on M series laptops. There was a clear case for it back when Apple silicon absolutely dominated anything else you could buy but with Panther Lake I'm not so sure that Apple's lead will continue to be big enough to justify needing to wait several years for full Linux hardware compatibility to be reverse engineered.

You can do pipelined option/result manipulation now, but without sum types/pattern matching it's not going to be quite as useful as Rust's.

Oh no the models are absolutely getting better, I'm just amazed that only 6th months ago I was using gpt-5.3-codex, and now I can use gpt-5.6-luna for similar results at like 1/15th the cost. Now 5.6-sol is being slashed by 50%? Amazing.

I'm loving this race to the bottom.

I like Go but a couple of these "advantages" wash out when you add the scale and typical usage patterns of agents.

| Go is Readable / Go is Maintainable

It's true that Go, as a low-magic language, tends to be very same-y looking across projects, which is incredible for being able to reliably understand your dependencies' source code. And its tooling is world-class. I love this about Go.

But in practice I've found that, working in a monorepo with multiple teams, contributors that don't have cross-team legibility as a priority will just write SO much more code. And with business logic, often the fact that I can read the code on a line-by-line level doesn't matter if I don't understand the wider context to know how something might effect spooky action at a distance.

Pre-agents, I witnessed a fast transition from a codebase that I could mostly hold in my head to one where large swathes of it had been written and rewritten until they were unrecognizable to me. Now we have agents and, since they are still mostly not good at software engineering in-the-large, the process of knowledge debt accumulation (and ofc tech debt accumulation) in a codebase accelerates tenfold without concerted effort in the other direction. Go being easy to read does not intrinsically help with that.