LarsDu88

5 ms·

1.
by LarsDu88 4d ago
The end of the Sengoku Jidai was triggered by a warlord (Nobunaga) who equipped his armies with massed ashigaru levies armed with long spears and tanegashima arquebuses. It was absolutely a technological brawl where the valiant highly train
2.
by LarsDu88 7d ago
I really hated the fact that Fallout 4 doesn't communicate to you what missions are procedural and which ones are actually handcrafted. This is the most frustrating part of that game. The procedural missions lack narrative closure, so
3.
by LarsDu88 7d ago
This is something I realized about game design: if you have a procedurally generated world like XCOM, almost every playthrough (or "dungeon") has to produce something new to the player. Something like 10% of each run should actual
4.
by LarsDu88 7d ago
Evidence? Are you Grok bot?
5.
by LarsDu88 8d ago
Where exactly are you getting this view that folks at Anthropic want to eradicate 9/10ths of the planet's population? Who exactly is pushing this viewpoint?
6.
by LarsDu88 8d ago
I'm not an expert in fluid dynamics, but does this result have any positive implications for nuclear fusion research?
7.
by LarsDu88 16d ago
This feels premature. The thing about the LLMs people use today is that it's a handful of super expensively trained models serving 1000s of use cases ranging from frontier math to recipe planning. We get to learn the foibles and langua
8.
by LarsDu88 20d ago
I loved these games as a kid. So much so that I made a VR clone of sorts Rogue Stargun ( https://roguestargun.com ) Never quite got around to polishing it to the desired level (due to having kids), and Star Wars Squadrons ended up
9.
by LarsDu88 20d ago
Good ol commoditize your compliment model. Sell GPUs? Make LLMs cheap. If Anthropic and OpenAI makes their own inference hardware, this is a shot across the bow for making their business worthless. Expect the next NEMOTRON to top the AI per
10.
by LarsDu88 22d ago
Well Sam Altman finally has built a moat against Chinese open weight AI. Well done. But what will this mean for Cerebras? I remember when Tesla was building its own inference chips, and after about 2 years and billions spent, the whole effo
11.
by LarsDu88 23d ago
He has, but I think he's right in the sense that there will be better models. It would be premature to think that the current frontier model architectures will last into the future. It would be like making DVD engineering a standard pa
12.
by LarsDu88 23d ago
I agree more with Yann LeCunn's salty reply. Over long run, knowing how autoregressive language models work from scratch will be just one step in having foundational understanding, and they might become dated... like knowing how a CRT
13.
by LarsDu88 25d ago
This explains the grifting. Sorry Cloutier.
14.
by LarsDu88 26d ago
Oh I'm sure it very enjoyable. This is more on the question of "could it have been a better game with better reviews" if more time had been spent on gamedev versus spacetimedb? I suspect the answer to that is probably nuanced
15.
by LarsDu88 27d ago
I remember a friend showing me some stuff about this when they were still trying to pursue the MMORPG route and I am surprised this company is still alive given the amount of unnecessary overengineering that went into that effort. The model
16.
by LarsDu88 27d ago
Ok, you are correct. These models don't train noise predictors at all unlike first gen image diffusion
17.
by LarsDu88 27d ago
Denoising is probably the weakest part of this model. There is recent research from Kaiming He showing that predicting the noise is actually not the best strategy for image generation since noise space is so large. Simply predicting the sur
18.
by LarsDu88 28d ago
I was thinking about this approach, and this paper basically validates it. The inference cost must be extremely high compared to diffusion based approaches. I wonder if this would ever be useful for more organic sculpting type workflows. E
19.
by LarsDu88 28d ago
A design that bakes the architecture into silicon would be 10x faster, and imagine a version that does all the multiplication ops using single log-amp addition versus dozens of transistors to cut down the amount of silicon used by 50x. The
20.
by LarsDu88 28d ago
That's a Lord of the Rings length book from a frontier model (Sol) every 15 minutes.
21.
by LarsDu88 29d ago
The fools thought we needed data centers for agents, but we actually needed data centers for Github!
22.
by LarsDu88 1mo ago
Bevy is very much an engine. It's an just an engine without an editor. A really, really well designed engine is just going to be a library of highly modular components with a nice editor or two thrown in. As the old adage goes "Bu
23.
by LarsDu88 1mo ago
This is the infamous "stack fallacy" https://techcrunch.com/2016/01/18/why-big-companies-keep-fai... The idea that the layer above your own is trivial to build (e.g. biology is just chemistry, or ch
24.
by LarsDu88 1mo ago
They finally did it. Someone decided to automate fun. Instead of having fun yourself, your AI agent can have fun for you.
25.
by LarsDu88 1mo ago
I'm sorry to say this, but the voxelization diffusion model I mentioned is effectively AI sculpting, and AI animation is already quite good (for humanoids at least) Virtually all the dominos have fallen except good topology at the mome
26.
by LarsDu88 1mo ago
The real algorithm that's needed is an intelligent AI mesh retopologizer. Honestly more of a reinforcement learning AI problem than a generative AI problem IMO
27.
by LarsDu88 1mo ago
The leading 3d gen techniques are latent diffusion models that work off voxelized intermediate representations of objects. When you convert these into meshes, very often they will have poor topology. Topology is important if you need to mod
28.
by LarsDu88 1mo ago
China is more authoritarian but the US under republican presidential administrations has been more more militarily aggressive. Every single Republican administration has initiated invasions of other countries and every Democratic administra
29.
by LarsDu88 1mo ago
Its not an old idea at all. Its from a 2012 article by Graham Alison followed by a 2017 book on the rise of China specifically that Xi Jinping read and referenced. Its claimed to be an old idea, but its actually only recently become popular
30.
by LarsDu88 1mo ago
RL can do behavior cloning, but really needs good simulations or verifiable environments to get to superhuman levels. That currently exists for math, coding, and a lot of videogames. Soon there will be good enough simulations for robotics.

More ›