One Step Closer To AGI

The beautiful irony here is watching people debate whether AI or humans are the real threat, while completely missing that the bell curve shows they're literally the same distribution . The top panel shows folks arguing about AI safety with the extremes thinking it's either totally controllable or apocalyptically dangerous. The bottom panel? Same exact curve, same exact percentages, just swap "AI" for "people." It's like running two identical unit tests but changing the variable name and being shocked they both pass. The 68% in the middle are just vibing with reasonable takes while the 0.1% tails are preparing bunkers or writing Medium articles about how everything is fine. The real kicker is that whoever made this probably used AI to generate it, creating a beautiful recursive loop of irony. Plot twist: maybe the dangerous ones are the 34% on each side who are slightly concerned but not enough to actually do anything about it. That's the sweet spot where bugs make it to production.

When NVIDIA keeps pushing DLSS to make games look so realistic you can count individual pores on character faces, but your GPU is already crying trying to run Cyberpunk at 60fps. The meme uses the "Guys, I don't want to be bread anymore" format but flips it - turns out hyper-realistic graphics are becoming too realistic and we're all starting to question if we actually need to see every individual hair follicle rendered in real-time. DLSS (Deep Learning Super Sampling) is NVIDIA's AI-powered upscaling tech that's supposed to make games run faster while looking better. But by version 5, we've apparently crossed into uncanny valley territory where games might start looking more real than reality itself. Maybe we peaked at DLSS 2 and should've just called it a day. Also, can we talk about how we went from "wow, look at those polygon counts!" to "please stop, I don't need photorealistic sweat droplets" in like two decades? Gaming has come full circle.

You spend months architecting the perfect solution with every port, protocol, and interface imaginable. Then Microsoft Copilot shows up like "hey bestie, let's chat about your feelings instead of actually solving anything." The gap between what developers want (actual tools that work) and what we get (another chatbot that'll suggest `npm install` for a hardware problem) has never been wider. At least the motherboard I/O panel won't gaslight you into thinking your USB-C port is "just a learning opportunity."

So someone submitted an AI-generated assembly patch to dav1d (a video decoder), and it was slower than C. Let that sink in. Assembly—the language you write when you want to squeeze every last CPU cycle out of your code—got outperformed by C because an AI wrote it. The entire point of hand-writing assembly is to achieve performance that compilers can't match. You're basically telling the compiler "step aside, I'll optimize this myself." But AI-generated assembly? That's like hiring a robot chef to make instant ramen and somehow ending up with something worse than the microwave version. Turns out AI doesn't understand cache lines, instruction pipelining, or the dark arts of SIMD optimization. It just vomits out syntactically correct assembly that runs like it's stuck in molasses. Modern C compilers have decades of optimization wizardry baked in—AI has... vibes.

CEO walks into a boardroom meeting: "We've dumped $1.2 trillion into AI R&D and nobody's buying. Solutions?" Team member 1: "Build more datacenters?" Team member 2: "Advertise harder?" Guy with actual brain cells: "Maybe stop investing in AI?" Out the window he goes. Because the tech industry's solution to AI not selling is obviously... more AI. It's like debugging by adding more print statements until your logs crash the server. The sunk cost fallacy has entered the chat, and it's bringing venture capital with it. Fun fact: Companies are literally spending billions to shove AI into products nobody asked for—your toaster doesn't need ChatGPT integration, Karen.

So the top downloaded free app right now is Claude, followed by ChatGPT and Google Gemini. Sandwiched between them at #3? DICK'S Sporting Goods. Because apparently when people aren't asking AI to debug their code or write their emails, they're shopping for sneakers and camping gear. The AI arms race has gotten so intense that three different LLMs are dominating the app store charts, but somehow a sporting goods retailer managed to wedge itself right in the middle. Maybe people need athletic equipment to physically run away from their AI-generated code suggestions. Or maybe they're just buying gear to touch grass after spending 12 hours arguing with Claude about TypeScript types. The real winner here is DICK'S marketing team, who somehow convinced people that shopping for workout clothes is more urgent than downloading Google's AI assistant.