Filligree

7 ms·

61.
by Filligree 4mo ago
Keep in mind these are two different things. Not all websites need to be accessible.
62.
by Filligree 4mo ago
> I'd argue back that LLMs likely have a better understanding of a11y conventions than I do as well. To make the obvious counterargument, “then you shouldn’t be creating websites at all”. I don’t actually believe this, but I know pe
63.
by Filligree 4mo ago
Sure, but nonexistence is the default. Has to be the default; most possible things do not exist. In this case we have no evidence for, so we stay with the default.
64.
by Filligree 4mo ago
It increases the chance of solder cracks, which is one major cause of failure. Thermal cycles in general will do that, however, and home computers are designed to survive a lot of them. So overall it’s not something I’d worry about.
65.
by Filligree 4mo ago
Is it a local LLM? Sibling seems to be assuming remote, but I have trouble imagining a TUI that inefficient.
66.
by Filligree 4mo ago
I feel like that says more about them. Feeling good about the existence of something, or bad about its absence, does not count as evidence.
67.
by Filligree 4mo ago
Doesn't that reduce to RJ45?
68.
by Filligree 4mo ago
How many PCIe lanes are you allocating? The card is obviously 16-lane, but it also has two ports; 40Gb total. In a server that’s fine, but if you want 10G in a desktop you’ll have a problem. I’m probably not telling you anything new. NICs u
69.
by Filligree 4mo ago
> I wonder how hard it would be to get an agent to send me a text message if it gets stuck on something. Not too hard, aka. I have a friend who did this. Rather than a text message, he uses IRC, but the effect should be the same. IRC is
70.
by Filligree 5mo ago
Or a Prusa. They even sell a variant with no network hardware at all, for the terminally paranoid.
71.
by Filligree 5mo ago
I mean. Yes? Probably?
72.
by Filligree 5mo ago
I can strongly recommend reading... anything else by Egan, just not that one. It's not that it's bad. The problem's the opposite: He poses an existentially dreadful question which I can't definitively answer with 'n
73.
by Filligree 5mo ago
The outcome wasn't guaranteed; that's the scary part. If Trump had decided to take a hand then it could have been drawn out for months at a minimum.
74.
by Filligree 5mo ago
I would say West is Europe, Japan and a few others. But I think I need a new word for that one.
75.
by Filligree 5mo ago
Print x twice. Not all “side effects” care about order. Better yet, define an order for parameter evaluation.
76.
by Filligree 5mo ago
Having run multiple 1.7 servers, the automatic ID allocation only sort of worked. It was by far the biggest headache.
77.
by Filligree 5mo ago
We’re certainly not dying out. At minimum, eventually civilisation would crack badly enough that we can no longer make contraceptives, nor other entertainment to distract from sex. Which I’m fairly sure is part of the picture. Evolution did
78.
by Filligree 5mo ago
The crisis isn’t one of total population, but one of having too many unproductive old people. There are solutions to that. One or more of them will eventually manifest, assuming we don’t outrun it by way of AI. Personally I’m betting on the
79.
by Filligree 5mo ago
Sure, Darwin always has the last laugh.
80.
by Filligree 5mo ago
The countries that do pay for it don’t have significantly higher fertility rates. I’d say there’s an elephant in the room: Childbirth sucks. If you want women to willingly subject themselves to that, you need either a culture that virtually
81.
by Filligree 5mo ago
The purpose of limiting memory use is so your computer does not become laggy as you run out of memory. We don’t do it for its own sake. But then, what’s the point in using an inherently laggy technique to save memory?
82.
by Filligree 5mo ago
"Multiple minutes" doesn't sound like a lot. With millions of shards, do you really need to regenerate the assignment layout every couple of minutes?
83.
Google says it likely thwarted effort by hacker group to use AI for 'mass(cnbc.com)
1 points by Filligree 5mo ago|1 comments
84.
by Filligree 5mo ago
Too coherent. You need to work on your simulation.
85.
by Filligree 5mo ago
There aren’t currently any alternatives which do an equally good job of printing, except for other Bambu products.
86.
by Filligree 5mo ago
A lot of things are toxic at high doses, yet have essentially no effect at low doses. The linear no-threshold model is a simplification that rarely matches reality.
87.
by Filligree 5mo ago
I’m aware. Unfortunately I’m an iPhone guy, and the software is more important than the hardware… Also I don’t need fully ruggedized. Just enough that a cover would be superfluous.
88.
by Filligree 5mo ago
Why would I need the compressed layers?
89.
by Filligree 5mo ago
I would prefer a phone that was robust enough to not need a cover, because covers add a great deal of size and weight. In the absence of such phones, I compromise on adding a cover.
90.
by Filligree 5mo ago
The last one. It would make sense to have a sandbox system, but they don’t.

More ›