jtrek

I remain surprised no Italian plumbers have visited high ranking Republicans. They're the worst people.

One should keep Chesterton's Fence in mind.

Idiots like Musk will see a system they don't understand and tear it down, and then people die.

I think one of the reasons some of the little kids in my life like me is I try to give them honest explanations. They don't always fully understand, but I think they appreciate getting answers. And probably appreciate the occasional "I don't know, actually. Let's look it up"

What is an example of a safety rule who’s conditions no longer exist which would not have the conditions almost immediately return if the rule was removed?

My job is populated by dinosaurs that only recently adopted git for version control. They had some rules and procedures that made a kind of sense when deployment meant "I'll scp the files to the prod server", but don't add value anymore.

Some people had a rule where after "deploy" they would SSH into prod and check the md5 hash of the files and compare them to their local copy. You don't have to do that.

They also wanted to only allow one person to work on a file at the same time because "you can overwrite their changes". Git handles that fine (unless you really fuck up the merge conflict, admittedly)

Sometimes the real reason is uncomfortable and they don't want to say it out loud. Like, "the CEO is an idiot, and wants it this way for stupid reasons"

Though maybe "the CEO doesn't understand how Google calendar works, so he thinks putting our time off in a shared spreadsheet is easier" would satisfy?

At my job a lot of stupid things come out of "someone high ranking doesn't understand computers" or "they don't benefit from fixing this, so it's easier for them to leave it stupid"

I find teams won't mark messages as read. Like I'll be focused on the chat and it'll stay unread. I have to click out and back into it.

When there's one of those horrible "teams" things with threads, clicking on a new message on the sidebar shows me just that message. No context. I have to click on the channel and find it.

The core problem is they have "chats" and then some other horrible thing that looks like channels but sucks. Maybe they wanted that to be like a message board? I hate it.

Most of my work uses "chats" instead, but that's horrible. No discoverability. A thousand permutations of people in chats. One for every meeting. (Was that important message in the standup chat? The planning chat? The side chat with the three competent guys?). And no threads.

Just give me channels with threads.

AI is giving stupid people a big force multiplier on their stupid.

If Sony was selling a backlog going back to 1980 at deep discount with no monthly fees and DRM-free at the seller's discretion, they would get less flak.

I think they have incentives to tag broadly but not accurately so they pick up more visitors. Tag it as "anal" when there's like a glimpse of butt, and you might get more views from people looking for that. There's no mechanism to punish inaccurate titles and tagging that I know of.

I'm pretty sure they don't invest in search because it's a poor return on investment. Most people just hit the first page, crank one out, and leave. The next cohort is similarly easily satisfied with their crap search. The people looking for specifics like "big tit short red hair reverse cowgirl pov outdoors at dusk" aren't profitable enough to spend engineering time on.

That's my hypothesis, anyway.

My hypothesis is that the least competent people use AI the most, and it's a big multiplier on their incompetence.

Despite civic ignorance, overwhelming majorities are grateful (86%) and proud (79%) to be Americans. Most also believe America is a land of opportunity (61%), and even more believe the American Dream (74%) is available to them personally.

big lol