tunesmith

7 ms·

61.
by tunesmith 9mo ago
parent poster is saying a healthy newsroom is much better than one guy. You're disagreeing by saying one guy is not better than an unhealthy newsroom.
62.
by tunesmith 9mo ago
cultural problem too... like even before AI in recent years there's been more of a societal push that it's fair game to just lie to people. Not that it didn't always happen, but it's more shameless now. Like... I don&#x2
63.
by tunesmith 9mo ago
I think the real distinction is whether the output came from the artist's human intention, or whether someone just said "let's just see what happens!"... it's sort of impossible to reach inside the artist's bra
64.
by tunesmith 10mo ago
As always, this requires nuance. Just yesterday and today, I did exactly that to my direct reports (I'm director-level). We had gotten a bug report, and the team had collectively looked into it and believed it was not our problem, but
65.
by tunesmith 10mo ago
So, I've been playing with an mcp server of my own... the api the mcp talks to is something that can create/edit/delete argument structures, like argument graphs - premises, lemmas, and conclusions. The server has a good synt
66.
by tunesmith 10mo ago
Why would you want to describe it as "loses to" rather than "beats"? People want to win, not lose.
67.
by tunesmith 11mo ago
Sometimes I think I'm the only person on the planet that grew up knowing it as "Rock Scissors Paper". I still defend that as the best name in English. Rock needs to go first. Rock beats Scissors beats Paper. It's the mos
68.
by tunesmith 11mo ago
I've been dealing with Codex CLI for a while and I love it, but I'm wondering if my thinking is just limited. While I'm starting discussions and creating plan docs, I've never been able to ask it to do anything that take
69.
by tunesmith 1y ago
Wonder how it compares with elkjs? I had written a few things that wrapped dot, which felt clunky, but have since had good success writing react components that use elkjs. Elkjs has a stupid amount of configuration options, some of which ar
70.
by tunesmith 1y ago
I went to a bar a few years ago and there was someone set up at a table in the corner with a weird-looking deck of cards and a sign that said, "help me test my game design!" That particular approach worked really well, but I think
71.
by tunesmith 1y ago
Data point: I run a site where users submit a record. There was a request months ago to allow users to edit the record after submitting. I put it off because while it's an established pattern it touches a lot of things and I found it a
72.
by tunesmith 1y ago
Too bad it's misogynistic. I'm not sure you already knew that. If I were rude enough to call you a name, I wonder what term I could use that would work either way!
73.
by tunesmith 1y ago
For those who don't know, OpenAI Codex CLI will now work with your ChatGPT plus or pro account. They barely announced it but it's on their github page. You don't have to use an api key.
74.
by tunesmith 1y ago
I felt like it was getting somewhere and then it pivoted to the stupid graph thing, which I can't seem to escape. Anyway, I think it'll be really interesting to see how this settles out over the next few weeks, and how that'l
75.
by tunesmith 1y ago
The 2h 15m is the length of tasks the model can complete with 50% probability. So longer is better in that sense. Or at least, "more advanced" and potentially "more dangerous".
76.
by tunesmith 1y ago
What's the thing you should replace Jenkins with at scale?
77.
by tunesmith 1y ago
With a test, it might link up some functionality with "why" and pass, but then what happens if a business requirement just isn't a requirement anymore? The test will still pass. I'm thinking of something sillier, like a
78.
by tunesmith 1y ago
They should invent a programming language that only compiles if the why is still true.
79.
by tunesmith 1y ago
Not the authorities. The sources.
80.
by tunesmith 1y ago
Yeah this seems like the specification/implementation problem. One can perfectly implement a bad spec, but coming up with the perfect spec is a human problem.
81.
by tunesmith 1y ago
It's just that they take turns. The top rivals in darts, or pool, or any other turn-taking game is the same way. They have to react to what their rival left them.
82.
by tunesmith 1y ago
I tend to like the ethos/logos/pathos model. Arguments from clever people can sound convincing because ethos gets mixed in. And anyone can temporarily confuse someone by using pathos. This is why it's better to have arguments
83.
by tunesmith 1y ago
The whole point is to make many such predictions and experience many outcomes. The goal is for your 70% predictions to be correct 70% of the time. We all have a gap between how confident we are and how often we're correct. Calibration,
84.
by tunesmith 1y ago
Would o3 pro be the first one that can reliably understand a gigantic congressional bill, to the point where it could analyze and warn of side effects?
85.
by tunesmith 1y ago
Weird article, it mentions three apps at the beginning, doesn't even test two of them in favor of two others it didn't mention, and then declares all the apps failures.
86.
by tunesmith 1y ago
On the other hand, Shinkolobwe didn't arrive by meteor - it wasn't there one day after not being there the day before.
87.
by tunesmith 1y ago
That's exactly the right level of pedantry. ;-) Right, I'd argue all of that is in the human ingenuity category. But in contrast, if we knew that some invention like grinding lenses or whatever was completely impossible until so
88.
by tunesmith 1y ago
One of my favorite things to wonder about is if our historical tech tree has ever been truly limited by anything other than human ingenuity. Like, if not for the random placement of smart people in the right places, who's to say we mig
89.
by tunesmith 1y ago
Are you saying Aider isn't as good as Cursor, or that Cursor isn't as good as Aider?
90.
by tunesmith 1y ago
Reading this, I'm struck with a familiar feeling. I wonder if other dem-leaning readers can identify. I remember the old joke about how economists have successfully predicted five of the last three recessions. Now it's more intens

More ›