I know many of us are afforded the luxury of being able to expense certain things for work. Having recently been tasked with researching effective software to use, it had me wondering what the biggest misuse of a corporate card might have been.

I have heard of people with years and years of misused funds that only get caught once they have been audited. Would love to hear your horror stories - or bragging rights depending on which way you look at it.

Also would very much appreciate any advice around software you'd recommend! The simpler the better

https://fortune.com/2025/07/11/indeed-glassdoor-layoffs-jobs-recruit-holdings-hisayuki-deko-idekoba-ai/

Hey everyone.... I’ve been spending the last few months learning how to monetize simple skills using just my phone and WiFi. It started with curiosity, a few sleepless nights, and a lot of trial and error but now I’ve made a bit of money using free tools like Canva, Notion, Gumroad, and Reddit itself. Recently I realized that we often overlook the smallest skills that could make us money if we leaned into them more: things like creating Notion templates, writing product descriptions, organizing info, or just knowing what to Google. So here’s what I’m curious about....What’s one “small” skill you learned or practiced that ended up helping you make actual money even if it wasn’t sexy or glamorous? Whether it’s flipping items, setting up automation, editing something for someone, or something niche... I’d love to hear. Let’s build a thread that helps people see what skills are really working out here 🙏

Finally started my own business of handmade bookmarks - Get a set of 3 handcrafted bookmarks for just : Rs - 230 Shipping included across India ✨

I’ve always had this grand idea to open a laundry mat and I have a lot of ideas for vending machine items but I also thought about laundry folding services on the side. I would love to hear how well this type of business actually performs.

Mainly about overall business, obviously not looking for anything specific to what I want to do

AWS says Rainier strings hundreds of thousands of custom Trainium 2 chips into one continent-spanning UltraCluster, promising five-times Anthropic’s current Claude compute and 30-40 % better price/performance than its top GPU fleet. Add in a water-use target of 0.15 L / kWh and you’ve got a green-leaning, vertically integrated answer to the Nvidia supply crunch.

For business leaders: If months-long model training collapses to days and compute costs fall, the old bottlenecks move upstream - data quality, evaluation, and deployment speed. Does this change your 2025 AI roadmap or cap-ex plans?

Curious how CFOs, CTOs, and founders here see the knock-on effects - especially for SMEs and growth-stage startups that couldn’t outbid Big Tech for GPUs.