Prentis, a new AI lab co-founded by LinkedIn's Reid Hoffman, is in talks to raise about $100 million, according to TechCrunch. Its bet is simple. The biggest use of AI will not be writing code, but automating everyday computer tasks. Handling email, making documents, filling in forms — the dull work almost everyone does daily — could become the main battleground. For years, AI investment focused on code generation for engineers. Now that focus is shifting toward general office work.

Behind this bet is AI's growing ability to "see and operate" a screen. Older AI was good at producing text but weak at using apps for you. Recently, "agent" technology that recognizes buttons and fields and acts step by step has become more realistic. This is the base that makes everyday automation possible. From a long-term view, the key question is who will own this space. Code generation is already crowded with strong rivals. But automating dull office work has no clear winner yet. Backing this area now looks like an early move into open ground.

This wave of automation runs on data centers working at full power. Yet TechCrunch reports that when a single power line fell in Northern Virginia, it exposed a serious weakness in AI data centers. The centers did not respond well to grid disruptions, which could throw supply and demand off balance. AI performance is not decided by models alone. Power, grids, and cooling — the physical infrastructure — set a quiet ceiling. For your long-term finances, it is worth watching not only the flashy side of AI but also the plain infrastructure that supports it.

At the same time, pushback is growing beneath the adoption. According to TechCrunch, "Avoiding AI" workshops held at libraries across the U.S. are drawing unprecedented demand. People tired of Big Tech's all-AI push are choosing to keep their distance on purpose. Every adoption curve has a group that resists. This does not mean automation will stall. But it suggests that for AI services to be accepted, they need trust and comfort, not just performance. The more a product automates daily work, the more this acceptance decides its spread.

In short, AI's center of gravity is moving from code generation for experts to everyday work for everyone. The entry of well-known founders shows the bet is serious. Still, two walls stand in the way: the physical limit of power infrastructure, and the mental resistance of AI fatigue. For your own job, a good first step is to list which parts of your work are routine. The more routine a task is, the easier it is to automate. Going forward, it is wise to watch calmly how reliably these agent products actually perform at work.