🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

everyone else has meet-cutes for their relationships and crushes, and I'm just over here like: "idk he ranted about Fourier transforms" "he made me coffee at a deep learning thing" "he kept beating me at Scrabble" "he tricked me into creating a physics engine for his web app"

you ask people in Colorado “so what do you do” and they just say an outdoor sport legit don’t think I’ve learned what anyone does here for work

After a successful two-year experiment, we’re making it official. Lyft is a fully flexible workplace. Work from the office, work from home, work wherever you work best. We’re also hiring. A lot. Join us: lyft.com/careers

This is a damn trap

okay but why are streaming services so obsessed with mid-2010s tech disaster stories all of a sudden

One of the most valuable things you can do in analysis-focused data work is to establish a consistent set of baseline definitions and facts, then making sure they are available to other folks and easy for them to incorporate into whatever they're doing

You cannot get your confidence from other people. You have to know who You are.

cardinal rule of working in big tech: do not update macOS until instructed to

wow, the devil works hard but these Hollywood actors publicists work hard

ask not what your country can do for you. they literally won’t do it

the devil works hard but the algorithm immediately targeting me with pride ads on june 1 works harder

Sayash Kapoor The hard part isn't building the model—it's assembling the data, evaluating performance, integrating it into existing workflows, and adding oversight. All those are site specific and manual. Buying a tool instead of building the model in-house only makes them harder!

Yes, ChatGPT is amazing and impressive. No, OpenAI has not come close to addressing the problem of bias. Filters appear to be bypassed with simple tricks, and superficially masked. And what is lurking inside is egregious. Abeba Birhane Sam Altman tw racism, sexism.

When the output is beautiful, it's all the model. When the output is racist, discriminatory, or otherwise unjust, it's the training data, the filtering, or society

At this time next week we'll be knee-deep in stories about the trials and triumphs of working with data. Can't wait! 🍌 Register for Data Mishaps Night (and check out the lineup!!) here: datamishapsnight.com

Basically without data and model transparency, open-sourcing "the algorithm" doesn't mean much.

A brutal reminder from The Markup that while we can worry/hope about the effects of LLMs, there's AI used TODAY in consequential decisions is affecting our lives RIGHT NOW. themarkup.org/the-breakdown/…

someone deploy Michael Lewis on this immediately

The mishap is coming from within the house 😂 a great #datamishapsnight

|￣￣￣￣￣￣￣￣￣￣￣￣￣￣￣￣￣| | Don't Push To Production On Friday | |＿＿＿＿＿＿＿＿＿＿＿＿＿＿＿＿＿| \ (•◡•) / \ / —— | | |_ |_