“Hi! I’m Your AI and I’m Here to Help.”

The pendulum of the AI automate/augment debate continues to swing. For a while the buzz around reasoning agents delivering better-than-human performance (prematurely) assumed that AI would replace entire job categories and the imminent unemployment and macroeconomic earthquake captured everyone’s attention. More recently, with the disappointment around GPT5 (what? no AGI??) and the narrow MIT study about enterprise AI project success (which proved the old adage that “he who aims at nothing will certainly hit it”), the bob has swung back into augment territory with rising interest in task specific agents.

The truth is, of course, both, but also everything in between as well.

There is a false dichotomy in AI conversations: automation vs. augmentation. The old patterns are holding us back. What if AI, instead of elevating the bottom performers to average, elevated the top performers to superhuman. Which one impacts the bottom line more? Which one is more likely to make someone’s career?

Reality, as the wise man said, is messy. Reasoning agents show promise, but we’re not quite there yet. On the other hand, task agents are here today and delivering tons of value. The path forward is a hybrid. Automate what you can. Augment everything else.

The key insight is to stop thinking about reasoning agents (at least in the short term) as taking on roles that people inhabit. Instead, we should be focusing on the pragmatic use of task agents to automate specific activities with humans as process designer, coach and orchestrator.

Here's what life with agents actually looks like.

Instead of thinking "how do I solve this problem," professionals start thinking "what agents do I need for this?" You build a solutions catalog in your mind. Market research agent for competitive analysis. Lead gen agent for prospecting. Data analysis agent for patterns.

Your core skills change. You get better at spotting the right problems to solve. You learn to set clearer goals. You manage teams of agents and people together.

Creating these agent teams becomes as easy as inviting people to a Google Doc. You find, connect, and deploy agents in minutes. And these teams of agents are burning exponentially increasing numbers of tokens over time - cramming days, weeks, even months of work into your coffee break or night off.

The real challenge is how do people keep up?

This speed creates something unexpected. Ambition. When you can set up the foundational work so quickly, you have time to go deeper and explore a wider range of possibilities than you ever could before.

Serendipity happens just outside what you already know or can see. Agents give you the bandwidth to go there.

The question isn't whether agents will change work (They will. They are.). The question is whether we'll use them to optimize the status quo or, aspirationally, use them to dream bigger. You know which way we’re leaning at Hubs.is.

Scott Wiener,  CEO and Founder at Hubs