Senior Staff Eng - Android, GraphQL, AI @wayfair . Views are my own. Quick learner. independent. 🏃🏻♂️🏂⛷🏓
Boston, MA
Joined September 2011
- Tweets6.2K
- Following429
- Followers341
- Likes1.5K
Honestly can’t wait until the day we can actually replace gradle. Might even be the day Gradle actually fixes their problems 😆
Overview of Lightbuild
developer.android.com/tools/…
NY Post in a nutshell
This. It’s just a sad reality many of us will have to accept.
One of the largest joys I got as I moved up in my career was my ability to listen to engineers - understand their problem and almost immediately have a knack for coming up with 2-3 possible solves for them. Then seeing them implement those changes and see it succeed.
This is over. We get large 100 file PRs without any of that discussion, that banter with recurring cast of characters and folks attending my office hours.
AI give so many opportunities but it really takes the humanity out of it
This will not be used to scam elderly people into giving them gift cards
Introducing Griffin, the first model to pass the video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
It’s the first Human Interaction Model (HIM).
Readers added context they thought people might want to know
The 48% figure and "video Turing test" claim are from Tavus's own study of 54 one-minute calls, not independently verified or using a standard protocol. Griffin-Lite leads NVIDIA's VideoFDB benchmark on their public leaderboard.
cellcog.ai/blog/tavus-gri…
research.nvidia.com/labs/amri/proj…
tech-ish.com/2026/10/02/tav…
Andrew Grosner 🇺🇦 retweeted
Bill Burr has said AI should replace CEOs, not their working employees: “Why don’t they make an AI CEO? They’re making way more [than] all the workers. You’re going to save way more money replacing one guy. Why don’t you do that? Because it’s not for us. It’s for them.”
What the hell. Is this for real?
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
Guys got like 6-12 months left anyways
BREAKING: President Trump has rejected calls for an AI slowdown after CEOs of AI tech titans expressed safety concerns that the technology poses an existential threat, per FT.
"We’re leading China in AI ... and, frankly, I want to keep it that way, because whoever wins AI, wins,” Trump said.
The AI arms race appears to be accelerating.
First mistake was using Gemini
This is the stupidest thing I’ve ever read 😂
Three hiking novices asked Gemini how to climb 14,179-ft Mount Shasta.
It packed them for an 8-hour day.
They left camp 3 a.m.
They were supposed to summit by 7 p.m.
Gemini said pack simple carbs, not fats, because fats “take too long to digest”
Horrible advice 😂
Phones died. Backup chargers died.
One person blew their knee out.
Night fell on them at 11,000 ft in Mud Creek Canyon, with no extra layers.
They survived but barely.
“We relied too much on AI rather than our own critical thinking.”
You think? 😂
Strong disagree. With proper jetpack compose skills, your app can become more performant. Our apps are the most performant they’ve ever been, approaching best in class. Fully jetpack compose.
XML requires more tokens and output, which matters a lot more now . Less code AI has to write, the more effective and cost efficient
Very controversial take: with AI models getting this good, as an early adopter of Jetpack Compose, I honestly think Views are now a far more performant and reliable way to build Android apps.
The problem Jetpack Compose solved was verbose code, but it introduced the complexity of recomposition. Now AI can handle most of that verbose code anyway.
It’s funny how perspectives change with time.
95% of PRs are now cloud Devin @cognition . Estimated 6x in PRs a week now vs previously using primarily local agents . Crazy times!
AI slop
Celebrating small wins is hardwired into "Achieve," our goal-tracking sample app.
We wanted the simple act of checking off a task to feel like a celebration. Instead of the standard #Material ripple, we used Styles to drop in a custom, high-contrast completion animation—specifically, a brand-colored ring that expands and fades on tap, paired with a subtle pressed-state gradient.
It lets us focus entirely on polishing those custom, rewarding micro-interactions while Material handles the foundation.
#GoogleDesign #MaterialDesign #JetpackCompose #AndroidDev #UXDesign #InteractionDesign
Crazy how Gemini is not even in the top 10 yet has all of the potential training data on Android development
Android Bench evaluates LLMs against Android tasks to give you insight into how helpful models are for Android development.
We just added eight new models, including:
🔹 GPT 5.6 Sol
🔹 Claude Opus 5
🔹 Kimi K3
🔹 Qwen3.8 Max
Full leaderboard → goo.gle/4swdpRe
ALT A graphic announcing "8 New Models have been added to Android Bench" displayed over a stylized green bar chart with "AI" and sparkle icons.
