I actually did too lol. I was pleasantly surprised because it was actually decent and realistic about the situation (a lot of people get this romantic idea about going to the jungle to live and learn with the indigenous and have an "authentic" experience, and this does a pretty good job if dispelling that).
Heavy and chronic usage of cannabis is associated with some of these things. I think "plenty of evidence that marijuana can cause {effects}" is somewhat overstating the consensus on the topic.
Like many things in life, cannabis can be enjoyed responsibly or irresponsibly. Irresponsible use is inadvisable and can absolutely ruin your life and the lives of others around you. I see no issue with responsible use, though. All things in moderation. Alcohol, social media, and caffeine all come to mind as examples of other drugs that can be largely safe and enjoyable if used responsibly, but which become dangerous/harmful when moderation is abandoned.
I think some of the coolest changes in this release are on the nodes side of things — they added Closures (kinda like lambdas!), Bundles (tuples/structs, I guess?), and Repeat (loops!) (already was in Geo Nodes, now it's in Shader Nodes too).
Blender nodes have come a long way over the past decade and it's incredibly satisfying to see the care with which they have been developed. Blender's node editor is my personal favorite node editor I've ever used in any software, and I often find myself wishing other software adopted some of their UI and UX conventions.
Been a happy user since, oh, v2.75? And looking forward to being a user for many more releases to come.
Plus if you could use WASM modules as opposed to node systems, you would have a more powerful programming environment - for example, the mentioned "repeat" zone doesn't have "break" functionality, so you need to use a huge number of iterations and do an equivalent of "continue" once you're done, but even if you perfectly estimate the number of iterations you will need, the implementation is very slow. There are other problems like limited scope access, or in general slow node evaluation which make sequential algorithms (not reliant on parallelized work done by loops internal to many nodes) problematic.
Hello, my name is Kai and I'm a machine learning researcher with an interest in computer vision and novel neural network architecture design. I have 4 years of experience working on computer vision models for medical imagery, including publishing an echocardiography-focused finetune of CLIP in Nature Medicine last year. I'm self taught and enjoy work where I can be creative and learn new skills. I would love to be a part of the team that invents the Transformer-killer. I'm also open to more data-sciencey jobs if the work is interesting.
I also have 3D modeling, filmmaking/editing, and music-making skills and experience if they're relevant. I've been making movies and doing VFX ever since I was little.
My favorite Spherical Minecraft-like gamedev project is PlanetSmith [0], which uses hexagonal voxels (and a few pentagonal voxels). The devlogs are very well produced and I highly recommend checking them out.
> You can even reduce the amount of visible distortion by restricting players to a portion of a shell, so they never see the full difference in block size at its top and bottom.
> From what I can tell, this seems to be the approach used by the upcoming game PlanetSmith for its hexagonal-blocky planets.
My favorite way to generate random points on a n-dimensional sphere is to just sample n times from a Gaussian distribution to get a n-dimensional vector, and then normalizing that vector to the radius of the desired sphere.
Wonder if you get any numerical instability here in high dimensions by doing a sum of exponentials? Probably not because they’re Gaussian (no long tails) but after looking at scipy.special.logsumexp [1] I’m a bit wary of sums of exponentials with float32. Would be curious to see if there’s any characterization of this (the cited paper in the article only considers the low dimensional case)
This is exactly the method the article describes as the most common method (though the article uses the more specific “standard normal” rather than the more general “gaussian” when describing the distribution), and notes generalizes efficiently to higher dimensions unlike accept/reject, but, as the article notes, the accept/reject method is more efficient for n=3.
The only reason I can think of that you’re getting downvoted because this is mentioned in the article. This is a strictly better method than the accept/reject method for this application. The runtime of the accept reject algorithm is exponential in the dimension because the ratio between the volume of the sphere is exponentially smaller than the volume of the hypercube.
I’d also point out that the usual way Gaussians are generated (sample uniformly from the unit interval and remap via the Gaussian percentile) can be expressed as sampling from a d-dimensional cube and then post-processing as well, with the advantage that it works in one shot. (Edit: typo)
Have you kept up with recent ML papers like MindEye, which have managed to reconstruct seen images using image generator models conditioned on fMRI signals?
Ever since that paper came out, I (someone who works in ML but have no neuroimaging expertise) have been really excited for the future of noninvasive BCU.
Would also be curious to know if you have any thoughts on the several start-ups working in parallel on optimally pumped magnetometers for portable MEG helmets.
> Have you kept up with recent ML papers like MindEye, which have managed to reconstruct seen images using image generator models conditioned on fMRI signals?
Not really. I left the field mostly because I felt bitter. I find that most papers in the field are more engineering than research. I skimmed through the MindEye paper and don’t find it very interesting. It’s more of mapping of “people looking at images in a fMRI” to identifying the shown image. They make the leap of saying that this is usable to detect actual mind’s eye (they cite a paper where they requires 40 hours of per-subject training, on the specific dataset) which I quite doubt. Also we’re nowhere near having a portable fMRI.
As for portable MEG, assuming they can do it: it would be indeed interesting. Since it still relies on synchronized regions I don’t think high level thinking detection is possible but it could be better for detecting motor activity and some mental states.
reply