The AI Bubble Is BURSTING... Gazzettino [pK3ahxsLJLk]
Tag: #Gazzettino, #cristel carrisi, #jim jarvis, #margot robbie
Everyone keeps saying the same thing: project freedom AI is going to replace programmers.
Youve probably seen the demos AI generating apps, fixing bugs, writing entire features in seconds. It makes it look like coding is basically solved.
But a new benchmark tells a very different story.
Researchers created **SWE-Bench Pro**, a test designed to measure whether AI can actually perform **real software engineering work** inside large codebases. Instead of simple coding exercises, models had to explore real repositories, understand the system, fix bugs, and ensure nothing else broke.
The results were surprising.
But understanding large systems, debugging complex behavior, and designing software is still a completely different challenge.
If you're thinking about quitting celtic f.c. games coding because of AI, kneecap watch this first.
Subscribe for