#gamedev vet, C# fan, DOD/ECS expert. Made Robocraft & Svelto.ECS.
UK
Joined July 2009
- Tweets39K
- Following195
- Followers2K
- Likes12.5K
Pinned Tweet
I published the first video, without audio, of Polysmith, just to show a subset of the available feature. If you are interested, I will add explanations and I will make more video. I have already used it in a product, but I wonder if there is a general interest for similar tools
Sebastiano Mandalà retweeted
The renderer is there in 6.7 a lumen like indirect GI but only for difuse. I tested it for the client project. It is not bad.
Intel, Nvidia and Google totally collapsed...like memecoins...meanwhile my crypto portfolio is quite healthy
Two things Unity urgently needs:
URP and HDRP unification and a more realistic rendering, come on guys, with AI should not be a problem now
Full wasm32/64 multithreaded support!
Slowly people are realizing that the BIGGEST failure of AI is that people are still needed to write prompts.
There are several types of offensive culture appropriations 😂
bbc.co.uk/news/articles/cx4g…
At the end of the day all that matter is that after 10 years hating them, I was right about blueprints...
Disney lost creativity a long time ago, it's just about being squared and scared of the execs.
During the filming of Pirates of the Caribbean in 2002, Disney executives became deeply alarmed by Johnny Depp’s unique portrayal of Captain Jack Sparrow.
The studio’s top brass seriously feared that the actor’s eccentric mannerisms and peculiar way of speaking were completely ruining the film.
That bold, creative gamble ended up becoming a massive cultural phenomenon, establishing one of the most iconic and profitable characters in the history of modern cinema.
If you are interested in #robotics and what I am doing at Spinworks, check our web page, but most of all join our Discord! You can follow us to understand that what we are doing is not a game, it's democratizing robot training
robosim.spinworks.tech/
discord.com/invite/vxDREUUeW…
Already training the fence jumping, once done I am going to remove the code from the controller that detects heuristically what can be jumped (no markers used in this demo) and the policy will automatically detect any compatible situation if it works.
Adam is back! @unitygames
I see a lot of Motionbricks video and I wondered if dev understood that this tech is usable today. Unity supports it natively through Sentis package, so I wanted to show it off in a real setting.
No animation in this video is human made!
Adam is back! @unitygames
I see a lot of Motionbricks video and I wondered if dev understood that this tech is usable today. Unity supports it natively through Sentis package, so I wanted to show it off in a real setting.
No animation in this video is human made!
I cannot believe that it's 2026 and we have AI and I still have to use virtualdub to edit a video efficiently
it might not be a bad thing. The problem they want to solve is clear: abstract the user from the mess the current framework is.
AI fixing their own errors in text is so comical...They always first do the error and then to fix it they state the opposite in the text. Both statements were not needed in the first place!
Sebastiano Mandalà retweeted
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…