Well, there goes another one. Maybe Finance, the personal finance app that caught my attention with its sleek design and open-source promise, has just announced they’re shutting down their consumer-facing product to pivot to B2B. Their final version 0.6.0 dropped on GitHub with what I’d call a refreshingly honest explanation, but it still stings for anyone who bought into the vision.

This whole situation has me thinking about the fundamental tension between venture capital and open source software. When Maybe first appeared on my radar, something felt off about the setup. Here’s a company that raised VC money, promised an open-source personal finance tool, and then – surprise – discovered that giving away software for free doesn’t generate the returns their investors were expecting. Who could have seen that coming?

The tech world is buzzing with news of yet another groundbreaking open source AI model coming out of China - this time a 106B parameter Mixture of Experts (MoE) model that’s supposedly approaching GPT-4 levels of capability. And honestly, it’s got me thinking about how dramatically the landscape has shifted in just the past few months.

Remember when OpenAI was the undisputed king of the AI hill? When every major breakthrough seemed to come from Silicon Valley? Those days feel like ancient history now. Chinese companies like DeepSeek, Qwen, and now GLM are not just keeping pace - they’re setting the bloody pace. And they’re doing it all in the open, releasing their models for everyone to use, modify, and build upon.

Been following this whole Coldplay concert kiss cam drama that’s been doing the rounds online, and it’s got me thinking about something that goes way beyond celebrity gossip or infidelity scandals. The incident itself is almost beside the point - what really bothers me is this casual acceptance we seem to have developed around being filmed, broadcast, and potentially humiliated for the sake of “entertainment.”

The whole thing started when someone asked a pretty reasonable question about the legal implications of kiss cams and audience filming at concerts. They wondered about consent, about the fine print we never read, about whether we actually agree to having our most vulnerable moments broadcast to thousands of people. But instead of engaging with that important question, most of the discussion devolved into moral judgments about the people caught on camera.

There’s something deeply unsettling about watching democracy crumble in real-time from across the Pacific. The recent ruling that Trump’s firing of FTC commissioner Rebecca Slaughter was unlawful should feel like a victory for the rule of law, but honestly, it feels more like rearranging deck chairs on the Titanic.

The judge’s decision is clear: Trump violated protections for independent agency officials, and Slaughter remains a “rightful member” of the FTC. It’s the kind of ruling that would have meant something in, say, 2015. But we’re living in a different world now, one where “lawful” and “unlawful” have become increasingly meaningless terms when applied to those in power.

There’s something oddly satisfying about seeing a post from someone who’s just installed their first self-hosted doorbell camera. It takes me back to my own journey down the rabbit hole of home automation and self-hosted solutions. What starts as “I just want to see who’s at the door” quickly evolves into running virtual machines, comparing AI detection algorithms, and debating the merits of different RTSP streams at 2 AM.

The original poster mentioned they’re running MotionEye via Proxmox VE, which honestly brought a smile to my face. Here’s someone who’s taken the plunge into virtualisation just to avoid paying monthly fees to Ring or whoever. That’s the spirit that keeps the self-hosted community alive, even when it means wrestling with configuration files instead of just scanning a QR code.

The tech world loves a good underdog story, and this week delivered one in spades. OpenAI, the company that’s been positioning itself as the undisputed champion of artificial intelligence, was apparently set to release what they called a “state-of-the-art open source model.” Then Kimi dropped their K2 model, and suddenly OpenAI went quiet. The timing couldn’t have been more perfect - or more telling.

It’s hard not to see this as a microcosm of what’s happening in the AI space right now. The established players, flush with venture capital and billion-dollar valuations, are getting outmaneuvered by nimble competitors who aren’t weighed down by the same expectations and corporate bureaucracy. Someone in the discussion thread put it perfectly: “OAI: ‘Guys we’re releasing an open-source SOTA model, get ready gonna be epic, we’re so back!’ Kimi-K2: drops OAI: ‘jk’”

The Fair Work Commission’s recent decision to reinstate a worker who was sacked for accidentally sending an email to all staff has got me thinking about just how backwards our workplace priorities have become. The fact that this happened at Bravus (formerly Adani) somehow makes it even more infuriating, but that’s probably a rant for another day.

What really gets under my skin is how we’ve created this culture where a simple human error – something that happens to literally everyone who’s ever worked in an office – can cost someone their livelihood. Meanwhile, the company that couldn’t be bothered to implement basic email security measures gets to act like the victim.

A mate dropped a bombshell on me over coffee yesterday that’s been rattling around in my head ever since. Their sibling managed to rack up a $60,000 credit card debt playing Candy Crush. Let that sink in for a moment - sixty thousand dollars on a “free” mobile game.

This isn’t just about someone being financially irresponsible. This is about a system specifically designed to exploit vulnerable people, and it’s working exactly as intended.

Been scrolling through some discussions about Runway’s latest AI video generation demo, and honestly, it’s got me feeling a bit unsettled. The tech is genuinely impressive – we’re talking about AI that can take motion capture data and slap convincing digital skins onto it in real-time. The finger tracking alone is something that would have been pure magic just a couple of years ago.

But here’s what’s really getting under my skin: we’re hurtling toward a world where distinguishing between what’s real and what’s generated is becoming increasingly difficult. One user mentioned the “uncanny valley” is still very much alive, and while that’s true today, another person pointed out that by this time next year, we might be dealing with something entirely different. The rate of improvement is genuinely staggering.

The story of two Michigan residents building their own fiber ISP has been bouncing around my head all week. It’s one of those tales that perfectly captures the frustration so many of us feel with the state of internet infrastructure, not just in rural America but right here in Australia too.

What strikes me most about this story isn’t just the technical achievement – though that’s impressive enough – but the sheer determination required to say “enough is enough” and actually do something about it. These folks didn’t just complain about slow speeds or poor service; they rolled up their sleeves and decided to become part of the solution.