stravant

5 ms·

31.
by stravant 2y ago
This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
32.
by stravant 2y ago
I did a few thousand lines of this. In particular it was interesting learning about D3D retained mode as I did that part. What a weird piece of rendering history. Worth a search if you haven't heard about it before: D3DRM.
33.
by stravant 2y ago
It's also possible it isn't AGI hard and all you need is the ability to experiment with code along with a bit of agentic behavior. An AI doesn't need embodiment, understanding of physics / nature, or a lot of other thing
34.
by stravant 2y ago
I disagree because AI only has to get good enough at doing a single thing: AI research. From there things will probably go very fast. Self driving cars can't design themselves, once AI gets good enough it can
35.
by stravant 2y ago
One of my favorite successes was getting an LLM to write me a program to graph how I subjectively feel the heat of steam coming off of the noodles I'm pouring the water out from as a function of the ambient temperature. I was wondering
36.
by stravant 2y ago
The join version is also certainly less readable for anyone who doesn't normally work with databases and is just dipping into the database handling code for some reason.
37.
by stravant 2y ago
More like he's finally done price anchoring and ready to make some actual deals.
38.
by stravant 2y ago
Such an invalid access of a very small address probably does result from a nullptr error: struct BigObject { char stuff[0x9c]; // random fields int field; } BigObject* object = nullptr; printf(&qu
39.
by stravant 3y ago
There are surely reasonable ways to smoke test changes to the extent that they would catch the issue that came up here. E.g.: Have a gauntlet of 20 moderate complexity questions with machine checkable characteristics in the answer. A couple
40.
by stravant 3y ago
> then generate a polygonal representation of what you see It's really not that surprising since, to be honest, meshes suck. They're pretty general graphs but to actually work nicely they have to have really specific topologica
41.
by stravant 3y ago
You know what people hate even more though? Paying for Windows.
42.
by stravant 3y ago
Probably 20% of _coin fares_. Most people pay with pre-purchased tickets or have a transit pass.
43.
by stravant 3y ago
You're looking for SDF fonts. Very simple at their core but you can incrementally do more work to increase the fidelity you get out of them.
44.
by stravant 3y ago
The data probably wouldn't look so clean if it were skewed. If Google were doing something interesting it probably wouldn't be skewed only by a little bit.
45.
by stravant 3y ago
You really don't think someone involved can come up with something?
46.
by stravant 3y ago
The eye doesn't have to fully work to be useful. Even being able to differentiate light / dark would likely be useful.
47.
by stravant 3y ago
If they were allowed to get significantly ahead that status quo would likely be disrupted pretty fast.
48.
by stravant 3y ago
Because labs can't build a virus from the ground up. They'd be working from an existing virus and only making the smallest of adjustments.
49.
by stravant 3y ago
The complexities don't multiply here. The complexities would only multiply if you had to sort for EACH of something. We're only doing a single sort. > sorting the points along one axis which already gives O(nlog(n)) You're
50.
Protecting customers with generative AI indemnification(cloud.google.com)
113 points by stravant 3y ago|62 comments
51.
by stravant 3y ago
Better yet require always the same default for boolean flags, so that it's easier to reason about lifecycle for them.
52.
by stravant 3y ago
I'm a big YAGNI proponent but format version numbers are not something I would dare to YAGNI.
53.
by stravant 3y ago
I would dare to say that the person your consider your best engineer probably only had ass on chair 4 hours a day and your don't even realize it.
54.
by stravant 3y ago
Putting a price on IPV4 would be great. Actual monetary incentive for people to finally make IPV6 work.
55.
by stravant 3y ago
Value proposition. What are you going to do, _not_ walk around with a phone in your pocket?
56.
by stravant 3y ago
The issue is that you typically need an unaccessible level of depth for new players to avoid the game getting "solved" by the data being open. WotC has a huge problem with this in Magic the Gathering for instance. They used to pub
57.
by stravant 3y ago
Uhh... Occam's Razor?? A bunch of long term reddit users who dumped reddit are here now because it's one of the best alternatives. But reddit users are addicts so they're still using reddit to some degree and cross-posting st
58.
by stravant 3y ago
It gets even better, those malware even dynamically generate replacement wallets, finding ones that have similar starting and ending characters to the target one so that the victim can't easily recognize that the wallet has been replac
59.
by stravant 3y ago
I would not be so quick to say that this is an "obvious" outcome though. You could apply the same argument to computer hardware: General purpose CPUs are incredibly wasteful compared to specialized hardware (see Bitcoin mining whe
60.
by stravant 3y ago
Paying for ChatGPT and CoPilot. I would say CoPilot saves me 30-50% of the time it would have taken me by hand depending on what kind of coding task I'm working on.

More ›