Can someone explain why Vizcom has no hype? It’s an ai sketch-to-render (2D and 3D) app that is mature and actually used by professionals in industrial design. None of their tweets has gotten any engagement, but everyday on twitter some random vaporware app gets 5000 likes
Ramiz Asadullaev presented a 3D Steam Deck model so realistic, he had to specifically clarify it was a render and not a reference photo.
Made with Plasticity, Substance 3D, and Marmoset Toolbag: 80.lv/articles/this-…
The tariffs and economic uncertainty have somewhat negatively impacted Plasticity sales. Not hugely but it’s noticeable. Are any other indie software companies seeing the same thing?
I used to underestimate Plasticity. It seemed like just another 3D editor — nice, but unnecessary. Now, it’s my best friend and go-to tool. Fast, intuitive, with features that work exactly the way I need. It doesn’t get in the way — it helps. What used to feel like routine now
I dunno, Gemini 2.5 Pro doesn't seem that impressive to me at coding tasks. I just had a very frustrating experience with it trying to make a Zebra stripe shader.
I’m really enjoying o3. It’s good at planning/architecture and code review. It seems a bit smarter than other models, but more importantly it’s fast and responds CONCISELY. It’s one of the few models that feels like you’re talking to a grownup
Just had Claude 4 do an "easy" but massive refactoring across 195 files. It required some babysitting but it worked. This couldn't have been done with simple grep/sed commands because many call-sites were unique. Via Claude code, total cost: $4.97
Dear @cursor - now that you have an extra $900 million dollars, can you please hire a few people to work on Electron ?? The current maintenance team is understaffed, underfunded, and there are lots of bug fixes and performance optimizations that would benefit practically the
Plasticity 2025.2 is HERE.
Plasticity 2025.2 introduces a broad set of new tools and improvements focused on surface modeling, analysis, and workflow efficiency. This release adds G0–G2 surface alignment via the new Align command (powered by xNURBS), real-time continuity and
I just read the full version of this report and my interpretation is that much of the slowdown is from using chat/agents to edit code. It's not from autocomplete. Still an interesting finding! But this confirms my priors: autocomplete is the primary productivity boost in AI. Chat
I used vscode for a few hours today and it's still the case their AI autocomplete is way inferior to cursor. I'm really surprised they haven't caught up by now.