Jhambi

301 posts

Jhambi banner
Jhambi

Jhambi

@TheJhambi

hello

Katılım Ağustos 2013
851 Takip Edilen615 Takipçiler
Sabitlenmiş Tweet
Jhambi
Jhambi@TheJhambi·
#cottonart Figurine fanart. I debated finishing this one after seeing her excellent official 3D model but I had this about 80% done when that debuted so I figured might as well. I like the design a lot. More pics on my artstation.
Jhambi tweet mediaJhambi tweet mediaJhambi tweet mediaJhambi tweet media
English
12
8
268
14.5K
Jhambi
Jhambi@TheJhambi·
@Ruru_Berryz Do you know if this will get a normal wide release? I keep reading that it's only a super limited no-preorders in person prize thing and not a normal figurine release.
English
0
0
0
210
Ruru | MoePop
Ruru | MoePop@Ruru_Berryz·
＼ SQUARE ENIX Kuji ／ A Prize | POP UP PARADE Kaine L Size 「NieR Game Series 15th Anniversary Lottery」 ⟢ Release Date : March 20th #NieR #ニーアオートマタ
Ruru | MoePop tweet mediaRuru | MoePop tweet mediaRuru | MoePop tweet media
English
1
13
159
5.4K
Jhambi
Jhambi@TheJhambi·
@emollick Do you think an automated AI researcher allowing for automated 'benchmaxing' across many more areas/tasks would lead to solving some of that or just an even more uneven jaggedness as you hit areas it has and hasn't been (over)trained specifically for?
English
0
0
0
88
Ethan Mollick
Ethan Mollick@emollick·
Jaggedness remains a key feature of LLMs & I have yet to see a clearly articulated argument about why it will disappear. A jagged general intelligence (not quite an oxymoron, as humans are too) still creates lots of bottlenecks that require people & slow many kinds of take-off.
English
55
25
435
42.1K
Jhambi
Jhambi@TheJhambi·
@littmath That was a really good read, thanks for posting it. It can be hard to find a balanced/realistic take on some of this stuff.
English
0
0
1
356
Daniel Litt
Daniel Litt@littmath·
Some thoughts on AI and mathematics, inspired by "First Proof."
Daniel Litt tweet media
English
48
200
1.1K
334.3K
Jhambi
Jhambi@TheJhambi·
@drivascos What you are describing at the end makes me think of the image gen models trained to recreate what the person is thinking of from real time brain scan/MRI data. Thoughts straight to real time holodeck interaction (genie) is maybe not so far off.
English
0
0
1
42
852話(hakoniwa)
852話(hakoniwa)@8co28·
wan v2v refine dance test WIP 溶けの改善は…いい…が… 綺麗にはなるが引きの瞳感とハイライト感がぜんぜん違うのがどうにかならんか　コレばっかりはキャラのwanのlora作んないと駄目か　調整限界かも dance vidu / music suno
日本語
2
6
125
30.4K
Jhambi retweetledi
Google DeepMind
Google DeepMind@GoogleDeepMind·
Self-improvement 🔄 SIMA 2 can teach itself new skills, learning through trial-and-error, based on feedback from Gemini. Getting better the more it plays – without additional human input.
Google DeepMind tweet media
English
12
56
885
255.3K
Jhambi retweetledi
1X
1X@1x_tech·
NEO The Home Robot Order Today
English
6.7K
9.8K
68.5K
68.9M
Jhambi
Jhambi@TheJhambi·
@GioNakpil Lately I wonder how much will end up being a quick shuffle between different packages. With Tripo, Rodin, Hunyuan, etc image to 3D getting fairly good, I feel like its likely to end up sketch-> AI img->decent AI 3D, refine in ZBrush, texture (AI aided?), then right to AI video.
English
0
0
0
190
Gio Nakpil
Gio Nakpil@GioNakpil·
I’m exploring how 3D input can guide AI generation using Firefly Boards. I’m seeing promising results. Here’s a thread on what this could mean for artists in film and games who want more control over AI tools in a very assistive way (1/7)👇
English
3
2
12
2.9K
Jhambi
Jhambi@TheJhambi·
@aghdasiI Do you think that at a certain point, they will have added/specifically trained enough new tasks to where it starts to generalize more broadly to unseen ones with less (or even no) new training?
English
0
0
0
30
Jhambi
Jhambi@TheJhambi·
@The3DNinja The part that melts my brain is that its not just making it 3D, its still vectorized with selectable lines etc from any angle. Super impressive.
English
0
0
2
93
Wes McDermott
Wes McDermott@The3DNinja·
Checking out Turntable in the latest Illustrator (Beta) and totally blown away. Spin vector art like it’s 3D. Absolute wizardry! Grab the latest beta version of Illustrator and give it a try. #Illustrator
English
10
3
43
4.3K
Jhambi
Jhambi@TheJhambi·
@holynski_ So cool, I'd love to see how it handles more paintings like this.
English
0
0
0
219
Aleksander Holynski
Aleksander Holynski@holynski_·
Another one. Already a powerful painting, but moving around it yourself gives a totally different feeling. Jacques Louis David's "The Death of Socrates" => #Genie3
English
137
302
2.7K
320.8K
Jhambi
Jhambi@TheJhambi·
This is nuts. Knowing that it can work from existing video is interesting too, from the standpoint of eventually being able to generate a custom version of an existing movie/show in real time or interactively hopping into a scene.
Jakob Bauer@jkbr_ai

Something we discovered by accident: what happens if we start Genie 3 from a video and a completely unrelated prompt? Turns out the model really, really wants to make it work, to the point where it emulates itself. The prompt in this one is about a trex on a tropical island.

English
0
0
0
273
Jhambi retweetledi
Google DeepMind
Google DeepMind@GoogleDeepMind·
What if you could not only watch a generated video, but explore it too? 🌐 Genie 3 is our groundbreaking world model that creates interactive, playable environments from a single text prompt. From photorealistic landscapes to fantasy realms, the possibilities are endless. 🧵
English
813
2.6K
13.4K
3.7M
Jhambi
Jhambi@TheJhambi·
@kiaran_ritchie Watching the specular reflection stretch and squish as it goes down the side of the rocket thing is neat. It's always interesting to see what these models do (and don't) get correct, even when dealing with stylized images/video output.
English
0
0
2
155
Kiaran Ritchie
Kiaran Ritchie@kiaran_ritchie·
Everybody has wondered what it would look like if you could step into a painting and walk around inside it. For the first time ever, generative models are giving us a glimpse of that. Imagine walking around a watercolor forest in first person; with perfect temporal continuity in the linework and shading. (midjourney)
English
4
1
34
2.1K
Jhambi retweetledi
Alexander Wei
Alexander Wei@alexwei_·
On IMO P6 (without going into too much detail about our setup), the model "knew" it didn't have a correct solution. The model knowing when it didn't know was one of the early signs of life that made us excited about the underlying research direction!
Daniel Litt@littmath

One piece of info that seems important to me in terms of forecasting usefulness of new AI models for mathematics: did the gold-medal-winning models, which did not solve IMO problem 6, submit incorrect answers for it? 🧵

English
75
159
1.7K
281.6K
Jhambi
Jhambi@TheJhambi·
@flngr Initially, but I think for stuff like construction work, making some that size or even a bit larger would be super beneficial. At 7-8ft it doesn't need a ladder to frame or set stuff on the 2nd floor/roof etc at normal room/story heights.
English
0
0
1
31
Julian Bilcke
Julian Bilcke@flngr·
If humanoid robots were 7 feet tall do you think people would freak out?
English
4
0
4
656
Jhambi
Jhambi@TheJhambi·
@kiaran_ritchie Great write up. I've wondered how well current LLMs would handle playing D&D, especially given extra scaffolding (like the one playing pokemon). I think the first game to nail well integrated but open LLM NPCs will forever change how people view dialogue in games. dynamic vs flat
English
1
0
1
83
Jhambi retweetledi
Brett Adcock
Brett Adcock@adcock_brett·
Is this working Dan?
English
454
685
8.7K
2.5M

Keşfet

@Ruru_Berryz @emollick @littmath @drivascos @8co28 @GioNakpil @The3DNinja @holynski_