grace77

7 ms·

1.
Show HN: Agent Runner – open-source agent harness to benchmark real coding(designarena.ai)
4 points by grace77 10mo ago|0 comments
2.
by grace77 1y ago
yes - the changes since Show HN have been builders ( https://www.designarena.ai/builder ), audio ( https://www.designarena.ai/audio ), video ( https://www.designarena.ai/diffusion ), and compare
3.
by grace77 1y ago
We keep our system prompts across the board as bare bones as possible: https://www.designarena.ai/system-prompts As for good game dev prompts, here's one from a user that made a pretty fun game: Make asteroids with 2
4.
by grace77 1y ago
Exactly - we think the ai design tools are in the equivalent of the 'uncanny valley' territory that a lot of the diffusion models were stuck in just 1-2 months ago; most average diffusion models are still in this local optimum, bu
5.
by grace77 1y ago
yes! to the second point, someone in our show HN proposed encouraging human designers to compete in submissions as well - we tried implementing this and found that, at least right now, LLMs are still so bad at design that asking a human to
6.
by grace77 1y ago
yes - we're trying to figure out why that is
7.
by grace77 1y ago
we're not training models or proxying human votes with models
8.
by grace77 1y ago
yes! LMArena recently started pushing "webdev" arena, but there was no explicit emphasis on design or aesthetics, just web-based content
9.
Launch HN: Design Arena (YC S25) – Head-to-head AI benchmark for aesthetics
74 points by grace77 1y ago|24 comments
10.
by grace77 1y ago
yes! sorry to ping this post late, but we're experimenting with something now - excited to update :)
11.
by grace77 1y ago
ayy great to hear!
12.
by grace77 1y ago
yes! we have a copy code and copy react code button on https://www.designarena.ai/play
13.
by grace77 1y ago
I wish—just added them back
14.
by grace77 1y ago
yes! that is the hope — /play is our first attempt at building out utility, would love your feedback and will ship hard to make it happen!
15.
by grace77 1y ago
As are we, seems like the natural next step
16.
by grace77 1y ago
thank you! posting now :)
17.
by grace77 1y ago
Yes — great point. We originally waited for all model responses and randomized the vote order, but that made it a very bad user experience -- some models, especially open-source ones, took over 4 minutes to respond, leading to a high voter
18.
by grace77 1y ago
wow this is a super interesting idea, and the team loves it — we'll fast follow-through and follow-up here when we add it, thanks for the suggestion!
19.
Show HN: DesignArena – crowdsourced benchmark for AI-generated UI/UX(designarena.ai)
89 points by grace77 1y ago|29 comments