seif (@seifos) header
seif (@seifos) profile picture

@backboardio / @mcmasteru they see what they’ve been told to see

toronto
Joined August 2019
Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)Photo by seif (@seifos)
i turned a @solana hardware crypto wallet into a one-button voice transcriber for my Mac using @superwhisper @neilsuperduper @bpjzy One Button -> 2 Modes, with everything transcribed locally. Demo 👇
3
6
221
I won 2 @unruggable_io wallets this weekend, and figured i’d reengineer one of them into something i’d use everyday. So I wiped it and flashed new firmware that lets me use one button as a push to talk, or a recorder to transcript my thoughts into a markdown file, for either obsidian or claude
1
1
31
cloud optional, you can also swap out voice models pretty easily. next steps: get a battery pack or a longer usb-c cable lol!
1
15
when you open your pr thinking you’ll get a ‘LGTM 👍’ but end up with 17 comments
46
get your ice capp blended twice and thank me later
1
70
is it just me or has linkedin nerfed my posts like crazy
1
1
133
McMaster students! @Backboardio is teaming up with @DeltaHacks for Deltahacks Lite next Thursday, Nov 12 Join us for a night of building and hacking, with free food, AI credits, and prizes. What’s on the agenda: - mini hackathon: build and pitch a project in 2.5 hours - live demo of Backboard Studio - Q&A with Backboard engineers and ambassadors 6-10 PM, Marinucci Entrepreneurial Bridge, 7th floor MCCD. Sign up link in the comments, can’t wait to see you all there!
1
1
4
106
Welcome 🙏 First 100
4
102
Speed or quality? Today, if you serve AI models, improving one means giving up some of the other. To make models fast enough for heavy traffic, teams compress them, and that costs some accuracy. Backboard removes that trade-off. We ran the full, uncompressed Qwen3.8-Flash-Next on Backboard against a compressed version of the same model on a popular serving engine. The test was @SemiAnalysis_ independent benchmark. The full model on Backboard: → Started answering 2.9x sooner → Served 24% more requests on the same GPUs You run the full model, get more out of the hardware you already pay for, and have your cake and eat it too. Try it with your own models and traffic: backboard.io/blog/qwen-flash…
1
1
9
15,709
seif retweeted
We brought Toronto to GTA. Built at LarpHacks in 6 hours. 🇨🇦
4
4
9
28
4,322
We brought Toronto to GTA. Built at LarpHacks in 6 hours. 🇨🇦
4
4
9
28
4,322
The foundation is Slop Toronto, built on the open-source Slop New York project. We built the GTA-style gameplay layer on top of their massive city simulation, integrating vehicles, characters, and interactions into real Toronto geography. Huge credit to @johnfewell and his incredible project, which allowed us to build this game in just 6 hours (instead of 6 days).
1
2
167
Me and @ElkhoulyYouseph built this at LarpHacks, where we won 1st place for Best Use of Render. A ridiculous idea that somehow became a playable game. Huge thanks to everyone who helped make it happen. This was one of the most fun things I've ever built. What should we add next? 🇨🇦
1
106
check off roles to apply in 1 click with grok bot and notipply
3
145
been using cursor again and man i lowkey missed it
2
4
284
It is insane to me that there are still devs using terminals to prompt their agents
924
126
207
6,561
1,282,586