David Levin's Avatar

David Levin

@diwlevin

Sometimes professor sometimes nvidia research scientist all-the-time loud mouth Opinions are my own and do not represent either of my employers

546
Followers
98
Following
345
Posts
14.11.2023
Joined
Posts Following

Latest posts by David Levin @diwlevin

I emailed papersadmin@siggraph.org to avoid potential conflicts of interest. Let's see what the response will be.

11.03.2026 20:40 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Every one of my 3 #SIGGRAPH 2026 has one review (of four) that gptzero detects as 100% AI generated (not ai polished, or mixed, but 100% generated). 2/3 of those AI tagged reviews argue to reject the paper.

10.03.2026 15:16 πŸ‘ 6 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

in 2/3 cases that AI review is arguing to reject the paper.

10.03.2026 15:12 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

It takes real non effort to not even have an llm do your review for you. A true connoisseur of the lazy arts.

09.03.2026 21:45 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

We lowered the bar so much it just doesn’t matter any more 😬

06.03.2026 18:19 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

SIGGRAPH reviews are out, and I’m always reminded how, since SIGGRAPH doesn’t make anonymized reviews or discussion public, there is not even a minor disincentive to writing poor reviews (often violating the reviewing policy itself). I’ve been asking for this for years. It’s never going to happen.

06.03.2026 15:03 πŸ‘ 4 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I think you mean the LLMs writing the reviews are dead wrong. Weird how so many reviews have bulleted lists now …

06.03.2026 14:58 πŸ‘ 3 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0
Post image

Can't believe my new website didn't make it. www.diwlevin.com

05.03.2026 16:18 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Home | David I.W. Levin

My website has previously been described as
"where the internet goes to die" and
"the virtual embodiment of a calling card scam" and
"detrimental to <my> career".

Despite these compliments, I felt it was time to make something slightly more useful: www.diwlevin.com

04.03.2026 22:55 πŸ‘ 8 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

Not sure I understand this, perhaps Claude wrote this reply

03.03.2026 00:02 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

That’s the thing, now it has some value because LLMs need such a big data corpus. That multiplicity is what lets them learn a distribution. Now we can argue about how much value each nugget is worth (a lot to a little) but value it has.

02.03.2026 23:30 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I think it might also be a legal violation of the licences (IANAL). It would be fantastic if various government entities would grow a spine and enforce punitive financial penalties for this sort of thing, that get funnelled back to creators. Obviously the code has value and people should be paid.

02.03.2026 18:08 πŸ‘ 4 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

the centralization of power is definitely a legit concern. I guess there major solution is don't use these tools. I'm wondering how you feel about a world where these coding agents are trained and run locally, rather than as a SAAS type model ?

02.03.2026 17:19 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

for me it largely replaced cutting and pasting code off of stack overflow with a more efficient workflow. There's not substitute for thoroughly checking and debugging, but it helps me prototype ideas much faster.

02.03.2026 17:00 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

you plAIyer

02.03.2026 16:50 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

so its a blossoming romance ? You two are spending an awful lot of time together though ....

02.03.2026 16:29 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

The man who hated robots finds love with a robot.

02.03.2026 16:24 πŸ‘ 3 πŸ” 0 πŸ’¬ 3 πŸ“Œ 0

oh yes, repost this plz :)

02.03.2026 16:12 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I've wanted more of these types of papers forever in graphics. Benchmarks and evaluation are crucial in any mature field to help establish state-of-the-art. SCA 2026 is a great place to do that.

02.03.2026 16:12 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

"This year ... we would like to encourage submissions focusing on experimental/benchmark evaluation of existing methods. ... such submissions will ... judged on their ... experimental/benchmark design and methodology, insights gained and benefit to the community at large."

02.03.2026 16:12 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
The 25th annual Symposium on Computer Animation (SCA) will take place in Barcelona, Spain, from July 8 to July 10, 2026.

The Symposium on Computer Animation is Coming!

BARCELONA, SPAIN. JULY 8 - 10, 2026!

Papers Abstract Deadline: April 10, 2026
Poster Deadline: June 1, 2026

Website: sca.graphics

Call for papers : computeranimation.org/instructions...

Barcelona > LA (Come on: en.wikipedia.org/wiki/Sagrada...)

02.03.2026 16:12 πŸ‘ 6 πŸ” 3 πŸ’¬ 1 πŸ“Œ 0

Whoah Bruno that’s way too many … I think you have a problem :)

10.02.2026 20:41 πŸ‘ 3 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

OMG such an honour for me !

10.02.2026 17:27 πŸ‘ 3 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Can you imagine being the CFO β€œthe robots cost a million each AND we still gotta buy the hats? β€œ

30.01.2026 15:03 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

The robot doesn’t even need the hat !

30.01.2026 14:34 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Post image

Reminds me of photo-ops where politicians visit a job site. Look they’re wearing a hard hat and holding a hammer, they understand !

30.01.2026 14:33 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Cynically I think all this is a weird teleopereration play … if you can have your workers remotely controlling your robot factory all of a sudden you can do a lot more outsourcing and avoid point of origin manufacturing restrictions.

29.01.2026 15:41 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Post image

Don’t worry spaceX is going to take all that IPO money and buy a million Optimus robots.

29.01.2026 15:39 πŸ‘ 3 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

β€œGuys, you know that doctor octopus, you know from Spider-Man.. like what if he was all robot !”

β€œBro that’s genius”

29.01.2026 15:35 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Abhishek Madan

My PhD student Abhishek Madan is getting ready to graduate and he’s looking for a postdoc or industry position. Checkout his website www.dgp.toronto.edu/~amadan/

He’s been an amazing student , basically does all his research by himself. Fantastic math and implementation acumen.

29.01.2026 15:33 πŸ‘ 7 πŸ” 5 πŸ’¬ 0 πŸ“Œ 0