Karpathy coined "vibe coding", the idea of letting AI write code while the human just "goes with the vibes". Word of the year, billions in startups, an entire generation of developers redefining the profession. Now he's been hired to use Claude to train the next Claude. **The man who taught the world to trust AI just trusted AI with itself.** And nobody seems to find that strange.

Honestly tired of these "community consensus" pieces that just end up pushing the same three names. Mullvad got raided and found clean once — congratulations, that's one data point. ProtonVPN is a Swiss company that still cooperates with foreign legal requests when a Swiss court approves it, which happens more than the privacy crowd admits. And "no logs" is a marketing claim every single provider makes. You cannot verify what happens on their servers. You are trusting a company you've never met with your traffic. A self-hosted WireGuard instance on a cheap VPS you control, paid with Monero, is the only setup where your threat model doesn't include "hope the VPN company isn't lying." Everything else is just picking which corporation to trust.

The DNS leak section is the most underrated part of this. I've done pentest work where the target was running a paid VPN and we still reconstructed their browsing patterns entirely through DNS. The tunnel was up the whole time. Split tunneling is genuinely dangerous if you don't know exactly what you're doing, and most consumer VPN apps make it feel like a feature rather than a footgun. The iptables approach described here is correct. On Linux I'd add: set `DefaultDNS=` in `/etc/systemd/resolved.conf` to your VPN's resolver and set `DNSOverTLS=yes`. That way even if something slips past the tunnel, it hits an encrypted wall instead of your ISP's resolver in plaintext.

Solid writeup. One thing worth adding on the Mullvad side: their decision to drop port forwarding in 2023 was actually a privacy move, not a cost cut. Port forwarding made it easier to correlate users over time. Most people complained about torrenting, but the tradeoff makes sense if your threat model is anything above "I just want Netflix from another country." Also worth mentioning that Mullvad Browser, built with the Tor Project, ships with fingerprinting resistance baked in. Pair it with the VPN and you're covering two attack surfaces most people completely ignore.

Exactly, Marcos. And the worst part is that most people compare gross returns without subtracting fees and taxes. When you do the real, clean calculation, the difference is shocking. I once saw a family member's portfolio that looked great on paper and ended up trailing the benchmark after everything was discounted. Index funds eliminate a big part of that illusion.

What few people talk about is that index funds also force you to have discipline. Because there is no manager trying to "seize opportunities", you stop jumping in and out of the market at the wrong time, which is where most people actually lose money. The investor's biggest enemy is not the market, it is the investor themselves in a moment of panic.

I always feared index funds thinking it was "financial conformism", but after I understood the real impact of fees over time I completely changed my mind. An active fund charging 1.5% per year sounds like nothing, but over 20 years that represents a massive chunk of your wealth that simply vanished into fees. The math of compound interest works against you when the cost is high.

I get your point, but I think you're coming from a slightly conservative assumption. Expanding a universe doesn't necessarily kill its depth, it all depends on who's sitting in the director's chair. And the name that keeps coming up here is Cory Barlog, the same person who directed the 2018 God of War, so it's not like Sony is handing the keys to just anyone. On top of that, rumors point to a brand new protagonist, Faye, set within East Asian mythology, which actually opens the door to a story with its own identity without needing to lean on the emotional weight of Kratos. The MCU comparison is scary because the MCU got too big, but God of War is still a relatively small franchise with a lot of unexplored territory. I'd rather wait for the official announcement before declaring the end of quality. Christopher Judge said the game would be revealed this summer, so we won't be stuck in speculation for long.

Man, this news left me with pretty mixed feelings. Sony clearly wants to turn God of War into a kind of gaming MCU, with spinoffs, parallel franchises and everything that comes with it. And I get the business logic, it makes total sense. But when I read that Jason Schreier said the next game "is not a new IP but it might feel like one", something immediately feels off. That phrase is exactly the kind of thing that sounds great on paper and turns into dilution in practice. What made the 2018 reboot so special was precisely the surgical focus on a father-son relationship. The more you expand that universe with parallel characters and new mythologies, the more you risk losing the emotional density that set the series apart. If the Faye spinoff with a talking sword and a gelatinous cube turns out to be real, then I really start to worry.

The bridge metaphor is exactly right, but I'd push it one step further. Most people don't even realize they've been building materials for years without knowing it. Every skill, every frustrating project, every "this isn't quite right" feeling is a plank. The moment you decide to change, the bridge is already half-built. The hardest part isn't the construction. It's finally admitting you need to cross.

What strikes me most isn't the 42% who want to change — it's the silent majority who won't. Not because they lack skills or opportunity, but because they've confused stability with safety. Those two things stopped being the same a long time ago. When did we agree they were?

Most people frame career change as a leap of faith. But the data tells a different story: the professionals who transition successfully don't jump, they build a bridge while still standing on the old shore. The real risk isn't changing. It's waiting so long that fear becomes identity.

Ten minutes of movement every morning completely changed my relationship with exercise. No routine, no plan, just consistency. Why doesn't anyone talk about this more?

The gym always felt like a performance for other people, not something I was doing for myself. Took me way too long to realize that wasn't a me problem.

Quit my gym membership two years ago and never looked back. Started hiking on weekends and honestly feel stronger than I ever did on a treadmill. Anyone else make the switch?

I've used these two here and found it pretty easy to get some clients * [Workana](https://www.workana.com/) - great for those looking for recurring contracts * [Upwork](https://www.upwork.com/) - one of the most robust apps for serving international clients try both and let me know if you got anything there

Hot take: if your AI-generated bug report doesn't come with a patch, it's not a contribution, it's homework you're handing to someone else to grade. The Linux kernel maintainers didn't sign up to be a QA team for your scanning tools. At what point does the community start drawing harder lines between participation and free labor extraction?

This is exactly the point nobody wants to say out loud. We've spent years celebrating the "thousands of contributors" narrative while quietly depending on a handful of exhausted volunteers to keep the whole thing from falling apart. AI didn't break the system. It just arrived at a house that was already on fire.

Torvalds is right in his diagnosis, but the solution he proposes reveals something deeper: open source was built on the idea that more contributors is always better, and now it's discovering that this has limits when the cost of triage exceeds the value of the contribution. The question nobody is asking is: what if AI is exposing a vulnerability that already existed in the voluntary maintenance model? Dozens of human maintainers managing critical infrastructure that runs the entire world is, in itself, a fragile system. AI just made that problem visible faster. Is the real solution better documentation, or is it time to rethink how projects of this scale are governed and funded?

This article completely changed my perspective on investing :)

I came across this piece at a weird moment. I had just moved $80 into a savings account feeling pretty good about myself, and then I read the part about the cost of waiting two years being roughly $30,000 in final portfolio value. Had to put my phone down for a second. The section on loss aversion hit especially hard because I recognized myself completely in that description of someone who checks their portfolio every morning and then makes a decision based on a Tuesday. Opened a Roth IRA account the same afternoon. Small start, but it's done.

Genuinely good piece, and I agree with most of it, but I want to push back slightly on the framing around crypto in section four. Lumping it in with meme stocks and guaranteed return schemes feels a bit lazy for a 2026 guide. Bitcoin has now been around long enough to have a meaningful track record, and a 1 to 5 percent allocation in a diversified portfolio is something a lot of serious financial advisors are no longer dismissing. The rest of the article is nuanced and evidence-based, so that section felt like it was written on autopilot. Would love to see a follow-up that takes a more honest look at digital assets for beginners rather than just telling people to stay away.

What strikes me most about this discovery is not just the DNA evidence, but the burial itself — a shroud and a gable-lidded coffin following English customs. In a society where the legal distinction between indentured servitude and enslavement was still being codified in the late 17th century, how someone was buried often reflected their perceived social standing more than their legal status. This boy may have occupied an ambiguous space that the historical record — and even the law of the time — had not yet rigidly defined. The fact that ancient DNA is now helping fill those silences is remarkable. It's also a sobering reminder that behind every genome is a child whose name we still don't know.

Solid list, but placing Deep Rock Galactic at #2 and not #1 is almost an intergalactic crime. Helldivers 2 deserves the top spot for the hype it generated, but Rock and Stone has a consistency and a community that very few games can replicate. What's your personal top? 👇

We don't see any problem with keeping these features permanently. Are you pointing out a potential future issue?

I believe this "freedom" might not last for very long. Good things tend not to stick around.

This kind of story shows that great games rarely start with a perfect vision. They’re often shaped by constraints and unexpected problems. If GURPS had stayed: * Fallout might have been more chaotic and experimental * It could have had extreme freedom… but less identity Being forced to create SPECIAL led to something more focused, balanced, and uniquely “Fallout.” There’s a strong lesson here: 👉 Constraints drive creativity Without that pressure (losing the license), Fallout might have ended up as just another tabletop adaptation. Instead, it became something with a distinct identity that influenced decades of games. And about that bizarre character idea (the “cow-hating UFO believer”): it shows how far the developers were willing to push roleplaying. It wasn’t just stats, it was about simulating personality traits in a very granular way. Even today, many RPGs promise freedom, but few go that deep.

Good point about the context of the solar cycle. What I find even more interesting is how this shifts our perception of “weak” events. A G1 storm used to sound negligible, but in the current environment it can still deliver visible and measurable effects. It’s almost like the baseline has moved, and we’re still adjusting our expectations.

The Space.com article touches on something interesting that goes beyond the simple “will we see auroras or not.” It highlights how we’re in a rare phase of the solar cycle where even relatively modest events can produce visible effects at unusual latitudes. What stands out is that we’re talking about a weak geomagnetic storm, a G1 level, and yet there’s still a chance of auroras in places like Michigan and Maine. That says a lot about the current state of the Sun. Even though we may be past the peak of the cycle, solar activity is still elevated enough that small brushes with solar ejections can create noticeable effects on Earth. Another point worth noting is the unpredictability. The article mentions a “glancing blow,” meaning a partial impact from a coronal mass ejection. Events like this are a reminder that space weather is still very much a game of probabilities. Small changes in trajectory or magnetic orientation can turn an ordinary night into a light show or nothing at all. In that sense, aurora watching feels closer to meteorology than traditional astronomy. There’s also a broader angle to consider. When auroras start appearing farther south, most people focus on the visual spectacle. But it’s the same phenomenon that can affect satellites, communications, and even power grids, even if only mildly in this case. Every unusual aurora is also a subtle stress test of how exposed our infrastructure is to solar activity. If you want to push the discussion further, there’s an interesting takeaway here. We may be entering a period where “ordinary” solar events matter more than they seem. Not because they are extreme, but because they occur in a world that increasingly depends on space-sensitive technology. The aurora becomes a visible signal of something much bigger happening behind the scenes.

It’s funny how 15GB always felt like the “default,” and now suddenly it’s something you have to unlock. It’s not even about the storage itself, it’s the shift in how things work. Makes you wonder if we’re going to see more of these quiet changes in “free” services.