ip26

9 ms·

61.
by ip26 6mo ago
all risk comes with some small chance of tragedy This part. I’m not going to assume what that person meant, but there’s always a few people about in these conversations lamenting that when they were six, they were certainly never hit
62.
by ip26 6mo ago
Totally onboard with some managed risk of injury being ok. Very much not onboard with the parents who trust their six year olds to face mortal peril alone and make good choices.
63.
by ip26 6mo ago
So what do you recommend instead, a dice roll?
64.
by ip26 6mo ago
Good when the damage is done, bad as a method of approving new chemicals.
65.
by ip26 6mo ago
No living breathing human deserves to be subjected to my level of overthinking, and vanishingly few share my fascination with my favorite topics.
66.
by ip26 7mo ago
It certainly killed a lot of my tinkering outside of work, but that's more a matter of when I'm already doing the thing for most of the day, even though I like it I don't always want to continue doing it for the rest of the
67.
by ip26 7mo ago
If that reduced trafficking, would it not be a worthwhile exchange?
68.
by ip26 7mo ago
If this was all unequivocally true and the end of the story, then no one would make commission, and yet some do.
69.
by ip26 7mo ago
I keep asking it questions, and as I dialogue about the problem, I walk right into the conclusion myself, classic rubber duck. Or occasionally it will say something back, and it’s like “of course! That’s exactly what I’ve been circling with
70.
by ip26 7mo ago
It’s often lamented that some employees have a difficult case to argue for their impact on the bottom line, and as a result probably get paid a lower fraction of their value to the business than other roles where the link is easy to measure
71.
by ip26 7mo ago
That’s the odd psychosis here. Everyone knew loc was a terrible measure. But perhaps the instinctual pull was always there, and now that you can generate halfway coherent tens of thousands of loc in hours, our sensibilities are overwhelmed.
72.
by ip26 7mo ago
No one should care about actual code ever again. It’s ephemeral. Caveat: it still works best in a codebase that is already good. So while any one line of code is ephemeral, how is the overall codebase trending? Towards a bramble, or towar
73.
by ip26 7mo ago
The most exciting thing for me is it has changed the cost structure of studying code for refinement. I’d never done half as much code profiling & experimenting before. Now that generating one-shot code is cheap, I can send the agent off
74.
by ip26 7mo ago
Maybe the right answer is “why not both”, but subagents can also be used for that problem. That is, when something isn’t going as expected, fork a subagent to solve the problem and return with the answer. It’s interesting to imagine a singl
75.
by ip26 7mo ago
One significant difference is that all programming languages, whether machine code or python, have always been a precise language for describing the desired computation or algorithm. Working with an AI agent means specifying what you want t
76.
by ip26 7mo ago
Did you happen to be between 15-25 years old when ‘98 was released?
77.
by ip26 8mo ago
I had a somewhat similar experience - as I recall, most beeps happened as a result of a few stop lights with too-short yellows (e.g. the light changes yellow and you, even though you are below the speed limit, either panic stop or run the r
78.
by ip26 8mo ago
Some degree of road safety depends on predictable behavior. I haven’t seen those videos, but suddenly executing a panic stop on the freeway for no good reason at all increases everyone’s risk, even if the car behind you is following at a sa
79.
by ip26 8mo ago
I’m excited and waiting for the team that shows with $20k in credits they can substantially speed up the generated code by improving clang!
80.
by ip26 9mo ago
If the long term demand disappears, there may not be anyone left for TSMC to collect from on those MPA. This somewhat undermines their utility as a security.
81.
by ip26 9mo ago
Seems to me there are potentially opportunities for greater returns to data gathering work as quality data can inform many more papers in the future. How that will work still needs to be brokered…
82.
by ip26 9mo ago
You know where my child learned how to swim? A small pool, progressing from a hands-on adult to an arms-reach adult to lifeguards nearby. They may eventually swim in the sea, but they will have countless hours practice by then - most of the
83.
by ip26 9mo ago
Maybe the analogy here is you can buy only one kind of gun. That gun can load any ammunition. It’s easy for children to get their hands on anything from bb’s to BMG50. The gun has parental controls to allow selecting which ammunition the gu
84.
by ip26 9mo ago
Peculiar artifact? Are you forgetting all the people who retired from mining coal, building cars, or running plumbing? It's not weird that seventy year olds would prefer to stop doing those things. I don't know that many people wh
85.
by ip26 9mo ago
One possible positive outcome of all this could be sending LLMs to clean up oceans of low value tech debt. Let the humans move fast, let the machines straighten out and tidy up. The ROI of doing this is weak because of how long it takes an
86.
by ip26 9mo ago
I have some hope for transpiling to become more commonplace. What would happen if you could write in Python, but trivially transpile to C++ and back?
87.
by ip26 9mo ago
If you have two lanes and want three lanes, you could build the third as public, or as toll. If you build as public, it comes out of taxes, such as the gas tax. If you don’t have enough public money, perhaps you increase the tax. If you bui
88.
by ip26 10mo ago
Yeah, it’s a dumb term, sorry. Top spec? Whatever it is. Certainly it’s fair to argue that the top spec will continue to grow year over year, like happens with long lived desktop CPU sockets. Framework is bearing this out! But that spec doe
89.
by ip26 10mo ago
It’s an extreme framing for the sake of thought experiment. More specifically, I believe for “upgradable” to be a meaningful sellable feature you need something like this: LaptopA costs more than low-spec LaptopB. But LaptopB can be user up
90.
by ip26 10mo ago
Upgradable to what ? The ability to upgrade is well and good, but suppose the “endgame” configuration of an upgradable laptop was worse than the very base model of a non-upgradable. Why would you care about upgradability then?

More ›