Software stocks endured a difficult week, marked by sharp declines despite a late Friday rebound, continuing a challenging year for the sector. I reached out to our tech columnist, Ali Barr, who has been tracking AI adoption and its financial implications for enterprise businesses since launching our Tech Memo newsletter. He’s one of the most insightful voices on the evolving economics of AI and software. Ali, you’ve been examining how AI is reshaping enterprise spending and business models. Do you think the recent selloff in software stocks was overblown? The traditional software business model has powered the tech industry for decades. Companies invest heavily upfront to build software, but once it’s created, distributing additional copies costs nearly nothing. This leads to revenue growing much faster than costs, resulting in high profit margins—something that has made giants like Microsoft so valuable. But AI is fundamentally challenging that model. If AI dramatically boosts employee productivity, companies may need fewer software subscriptions. As AI tools get smarter, businesses could replace existing software entirely with AI-driven workflows or even build custom software themselves using AI coding assistants. And if software companies themselves adopt AI, their operating costs could rise significantly—because training and running AI models is expensive. That means higher usage might not translate into higher profits. If software in the AI era becomes less profitable and grows more slowly, it makes sense that investor confidence could wane. The stock market is reacting to a real shift in expectations. The scale of investment is staggering. Just Google and Amazon are planning capital expenditures of nearly $400 billion in 2026. If this continues for a few more years, total spending could surpass $1 trillion. To justify that, these companies need to generate over $1 trillion in new revenue in the coming years. AI is powerful and transformative, but investors are struggling to see the path to that return. Even if new AI products are revolutionary, will businesses and consumers have the budget to buy them all? That’s a serious question. One possible outcome is that Big Tech operates with slimmer margins in the AI era—similar to the pressures now hitting software stocks. Who are the most interesting figures to watch in this transformation? I’m closely following Andrej Karpathy. Former director of AI at Tesla and a founding member of OpenAI, he now operates independently, making his insights more credible. He also coined the term “vibe coding,” which captures the shift toward intuitive, AI-assisted development. Then there’s Aditya Agarwal, who led product engineering at Facebook and served as CTO and VP of engineering at Dropbox. Recently, he used Claude to complete a coding task and was deeply affected. “I am filled with wonder and also a profound sadness,” he wrote on X. “We will never ever write code by hand again. It doesn't make any sense to do so. Something I was very good at is now free and abundant.” I’m usually cautious, but early 2026 feels like a turning point—marked by profound disruption, and yes, destruction of old ways of working. The future of software isn’t just evolving. It’s being rewritten.