kadushka

11 ms·

211.
by kadushka 2y ago
Have you tried o1-pro?
212.
by kadushka 2y ago
I considered this when I sent them my sample. For me the benefit outweighed the risk. I'm sure their DNA database have been sold or given to all kind of companies and agencies since then (many years ago). Asking them to delete my recor
213.
by kadushka 2y ago
I paid them a lot more than $119 to have my DNA sample.
214.
by kadushka 2y ago
I don’t get it – why should I care that they have my DNA info?
215.
by kadushka 2y ago
I’m guessing about 80%
216.
by kadushka 2y ago
Did they say somewhere it was not remote controlled? When I watched the demo I assumed it was, because otherwise Jensen would’ve probably mentioned it.
217.
by kadushka 2y ago
1. People will continue to answer questions and post about events and problems after 2025. Eventually LLMs themselves (inside robots) will be observing the world and reporting on anything interesting. 2. Best LLMs today answer questions bet
218.
by kadushka 2y ago
A human runs on ~100W, even when not doing anything useful. It's entirely plausible that 100W will be enough to run a future AGI level model.
219.
by kadushka 2y ago
An article like this gets published a few times a year, every year in the last 50 years. Some years it comes true!
220.
by kadushka 2y ago
I make more than 250k as an engineer in a small startup. The only reason to join a FAANG is to make much more than 250k. According to levels.fyi an average senior SDE at amazon makes $412k total comp.
221.
by kadushka 2y ago
3.5B/14k
222.
by kadushka 2y ago
What recession?
223.
by kadushka 2y ago
That's surprisingly low for Amazon.
224.
by kadushka 2y ago
Wow, thank you for the link to the code - I haven't seen it before - it contains a ton of useful tricks. Lots to learn from there.
225.
Voice of America funding is cut(cbsnews.com)
8 points by kadushka 2y ago|0 comments
226.
by kadushka 2y ago
You mean soul?
227.
by kadushka 2y ago
Which model are you training and on what dataset?
228.
by kadushka 2y ago
At the end of the day, all modalities boil down to patterns of electrical activity in your brain.
229.
by kadushka 2y ago
Most modern LLMs are multimodal.
230.
by kadushka 2y ago
I just entered your prompt into ChatGPT (4.5): Prompt: I'm trying to make a rope, made of only aubergines, not using any twine or string. What kind of knot would you recommend to tie together each aubergine with the next? I'm try
231.
by kadushka 2y ago
There’s not enough improvement over regular LLMs to motivate optimization effort. Recall that the original transformer was well received because it was fast and scalable compared to RNNs.
232.
by kadushka 2y ago
I don’t know which text diffusion models you’re talking about, the latest and greatest is this one: https://arxiv.org/abs/2502.09992 and it’s extremely slow – couple of orders of magnitude slower than a regular LLM, ma
233.
by kadushka 2y ago
No, we typically train any diffusion model on a single step (randomly chosen).
234.
by kadushka 2y ago
Have they published their code?
235.
by kadushka 2y ago
That's the real readiness test for these agents.
236.
by kadushka 2y ago
Most people have someone telling them what to do at work, and checking the output.
237.
by kadushka 2y ago
Care to provide some examples?
238.
by kadushka 2y ago
I’m complete opposite.
239.
by kadushka 2y ago
Deep drawing: https://en.m.wikipedia.org/wiki/Deep_drawing
240.
by kadushka 2y ago
Don’t you feel like you don’t belong there with people half your age?

More ›