Wayne Chi
@iamwaynechi
Followers
461
Following
821
Media
21
Statuses
367
CS Ph.D. at @SCSatCMU. Funded by @NDSEG Fellowship. Editor at https://t.co/kBygvj9Puy. @MSFTResearch over Summer
Santa Clara
Joined July 2013
What do developers 𝘳𝘦𝘢𝘭𝘭𝘺 think of AI coding assistants? In October, we launched @CopilotArena to collect user preferences on real dev workflows. After months of live service, we’re here to share our findings in our recent preprint. Here's what we have learned /🧵
Introducing Copilot Arena - Interactive coding evaluation in the wild. Our extension lets you test top models for free, right in VSCode. Let's vote and build the Copilot leaderboard! Download here: https://t.co/Zyc9iL3u9m Led by @iamwaynechi and @valeriechen_ at CMU. 1/🧵
2
36
161
People like making arenas, but don't realize the sheer volume of battles you need to have good error bars...
0
0
4
I feel like Claude aggressively adjusted their limits again... I'm hitting the cap every single day now in what feels like 10 or less prompts
0
0
1
⚔️PR Arena is LIVE⚔️ We’re kicking off with 5 frontier LLMs tackling GitHub issue resolution. Resolve issues for free & vote for your preferred fix! 👉Leaderboard & Setup Guide: https://t.co/S1Oe3xWOhc
1
4
6
I think cookie clicker trained me to love ai research
waking up and feeling excited to check your experiment jobs from last night is what peak ai researcher lifestyle looks like.
0
0
2
Witnessing some lady try to enter taiwan with her global entry card
0
0
0
Ok great I guess all I had to do was post on twitter and the model stops messing up.
0
0
2
Is it just me or did sonnet 4.5 completely break Claude code? It's literally unusable and running into errors constantly 😂
1
0
1
I thought this was about @godotengine at first...
Kieran Culkin and Jazz Charton at the Waiting for Godot Opening Night!
0
0
0
we just killed software engineering btw
1
0
4
We are excited to launch the ⚔️PR Arena⚔️ leaderboard! Full results will be revealed after a certain milestone of community votes. Fix your GitHub issues for free and vote for better fix! 👉Leaderboard & Setup Guide: https://t.co/S1Oe3xXm6K
1
9
23
Sharing our initial leaderboard and open data release for 🎶Music Arena⚔️! Music is subjective and multi-dimensional. A key goal of Music Arena is to provide insights beyond binary preferences! 🧵
2
20
87
This might be one of the most unflattering photos ever taken
The Microsoft Research Undergraduate Internship Program offers 12-week internships in our Redmond, NYC, or New England labs for rising juniors and seniors who are passionate about technology. Apply by October 6: https://t.co/dGgXydhIQ3
1
0
7
PR Arena is finally out! I've been using it a lot for editing my website, coding subtasks, etc. Try it out!
Introducing ⚔️PR Arena⚔️ - free AI coding agents to fix real GitHub issues. Claude Sonnet 4 vs Gemini 2.5 Pro… Who writes better pull requests? 👉 Install here: https://t.co/bk19LcnBVf Powered by @allhands_ai
0
0
9
Skill issue
If your parents were reasonably well-off, you are probably downwardly mobile in relative rank. A child of the upper middle class feels dispirited because they actually have to try in order to merely reproduce their parents' success. Much Twitter Discourse is downstream of this.
0
0
1
Willing to bet the same is true for humans
We've entered an era where AI can win IMO gold but can't figure out how to spend United miles & PlusPoints coherently.
0
0
3
Not out of choice. This plane won't fucking leave. It's been 3 hours.
1
0
5
BREAKING NEWS! Making and comparing music is fun 😊 Turns out getting two drastically different musical interpretations is an incredibly different and fun experience. Try Music Arena out and let us know what you think! https://t.co/iT3rN8RDfH
beta.music-arena.org
Excited to share our beta release of Music Arena, a live evaluation platform for state-of-the-art AI music generation models! 🎧 Listen to the latest models and 🗳️ vote for your favorite ⚔️ https://t.co/XyXvzlOMcH ⭐️ https://t.co/9f3mvseEyu 📜 https://t.co/HqTg5PH09O
0
0
9
Pretty sure there's research saying the exact opposite of this
Holy sh*t… nobody’s talking about JSON prompting but it’s the cheat code to unlock God-tier outputs from GPT-4, Claude, and Gemini. Here’s how it works (with copy-paste templates): 👇
0
0
3