I'm looking ahead to the next wave of open-weight models that are as efficient as DSv4 (which is really efficient), and have been heavily distilled on GLM 5.2 (which is trivial, given it is open weight)
I use it all the time through Fireworks. The normal version when I pay it myself and the fast one when company pays. It's really fast and I never get rate limited with my daily use.
This is so reasonable and such a low bar to clear. Excited for this as the outcomes for people are happier and easier lives, instead of a lifetime of regret for learning old/outdated concepts at a premium price from teachers who aren't commercially viable in their own fields.
So it's a laminar flow fogscreen with 3d tracking of digits/inputs? Is this limited to single-viewer perspective then?
99% sure this is not novel (https://www.mdpi.com/2414-4088/7/7/63), but a cool combination of several ideas including tracking the eyes of the user to render content and a more detailed input tracking system. The real innovation here seems to be using multiple projectors vs a single one to get closer to a point-cloud style display.
The demos are cool! Could see entertainment and retail companies buying these if they are assembled into a cabinet, but fogscreens must always be in a dark environment and suffer from poor contrast, inability to black out or provide opaque sections, and always look misty/wavy even with tons of ultrasonic foggers and the laminar correction. If the cabinet controls for these factors and is place in an area with controlled overhead and surrounding lighting it could work.
Also, it does NOT require all viewers to share the same viewing angle.
It's a TRUE volumetric display, with the main caveat being projector-axis streaking. That streaking artifact could likely be reduced by tightly controlling both the viewing area and projector placement.
Since it's laser-based, you could theoretically use an unlimited number of projectors to reduce the smearing effect, trading some contrast ratio and black levels for cleaner volumetric imagery. (the streaking would still be there but in a circle pattern instead of from a handful of sources). Over time laser projectors could get so small that is possible but not practically today.
The biggest limitation is that the individual fog particles aren't controllable. The next step is likely some form of selective blanking of the volumetric reflectors, in this case, the water droplets. This would still not solve for the issue of occlusion of the projector to intended reflectors, but still, this is neat. This feels like the next evolution of FogScreen: interactive, true 3D volumetric imagery with depth!
I'm really excited about the concert entertainment, interactive storytelling media, and commercial retail applications once this becomes a buyable variance-free product. Can see it working in all sizes and scales. Very cool technology, and a great team behind the prototype. I'll be watching this closely!
btw ya'll gotta work with Disney Eng, Universal Eng, Triotech or Sally Dark Rides on some integration into dark rides and interactive coasters once this is ready. Would absolutely rock in that implementation once the concert/media versions are ready
People are drilling out the LEDs on glasses so people don't know theyre recording. This stuff is sickening and the users dont even realize or care what creeps they are.
I agree - the purchase price barely clears BOM before assembly. But I like that they're not really overpromising much, the use-cases are pretty narrow still.
If not due to their articles, then what cause do you attribute that widespread outcome among traded, subsidiary, et al. corporations to? (I’m happy to be wrong but it’ll take some time to research this and I want to be sure to hear your opinion in the meantime.)
The CEO and other execs will get paid more if they increase profits (usually in the short term). But that's very different than your claim that the articles of incorporation force them to do it.
reply