Why ‘Purpose-Less’ Science Is What Ends Up Changing Your Business

Last week, Sally Kornbluth, president of MIT, spoke to a packed room about something that feels very far removed from the reality of an SME: funding for university research in the United States. Her message was clear: curiosity-driven science —the kind that doesn’t try to solve a specific problem, but to understand how things work— is under pressure.

And you may be thinking: what does this have to do with my business?

A lot more than it seems. Let me explain.

The paradox of ‘useless’ research

The artificial intelligence that today answers your emails, transcribes your calls, or scores your leads didn’t come out of nowhere. It also didn’t come out of a lab with a clear commercial goal.

It came from decades of research that, at the time, looked like an academic whim:

  • Neural networks were studied in the 80s and 90s without any obvious practical application.
  • Transformers (the architecture behind ChatGPT) were born in 2017 in a Google paper that very few outside academia read.
  • Reinforcement learning was explored through games like chess or Go before being applied to industrial robotics or logistics.

None of that was invented with the goal of automating the inbox of a law firm in Madrid. And yet, here we are.

When someone defends ‘curiosity-driven science,’ what they’re really defending is the engine that produces the tools you’ll use in 5, 10, or 20 years.

Why this matters a lot if you run an SME

Kornbluth wasn’t giving a philosophical speech. She was warning about something concrete: if funding for basic research weakens, the flow of innovation that reaches the market slows down.

And SMEs are the ones that depend on that flow the most. Why?

Type of company How it accesses innovation
Large corporation Has in-house R&D departments, buys startups, funds universities
SME Waits for technology to mature and become accessible and affordable

Large companies can afford to experiment with expensive, unproven technology. SMEs can’t. You adopt technology when it’s ready to deliver results without needing a team of PhDs to keep it running.

That is exactly what is happening now with AI applied to business processes.

The sweet spot: when science reaches your street

Generative AI, autonomous agents, language models capable of understanding complex instructions... all of that is already out of the lab. It’s available via API, can be integrated, is affordable, and —most importantly— reliable enough to automate real work.

This is what we call a technology’s sweet spot:

  1. The basic science has already been done (at universities, over decades).
  2. Big tech turned it into a product (OpenAI, Anthropic, Google).
  3. Now any business can use it without building anything from scratch.

The problem is that many SMEs don’t realize they’re at this point. They still think AI is ‘for the giants’ or that you need an internal technical team to take advantage of it.

That’s not true.

What today’s research lets you do tomorrow (in your company)

Let’s look at concrete examples. Things that five years ago were science fiction and today are automations working in real businesses:

Calls answered 24/7 by a bot that understands context. Before: hire a call center or miss calls after hours. Today: a bot trained on your business information picks up the phone, takes messages, and schedules meetings.

Automatic lead scoring. Before: a salesperson spent hours researching each contact on LinkedIn. Today: each lead is automatically enriched with public data and enters the CRM already prioritized.

An inbox on autopilot. Before: two or three hours a day classifying, replying to, and forwarding emails. Today: 15 minutes to review what the system has already filtered, answered, or escalated.

Sales proposals in minutes. Before: one or two days per proposal. Today: 10 minutes, tripling the team’s capacity.

None of this requires you to understand how a transformer works. Researchers already took care of that —the ones Kornbluth is defending.

The mistake many SMEs are making right now

There are three typical reactions to the current AI wave:

  • Ignore it. ‘This isn’t for my business.’ Mistake: your competition is probably already automating something.
  • Buy random tools. Subscribe to five different SaaS products hoping they’ll solve the problem. Result: more tools that nobody uses well.
  • Wait for the perfect product. ‘When the final, definitive solution comes out, I’ll buy it.’ Meanwhile, you keep losing four hours a day to repetitive tasks.

The fourth option —the one we recommend— is to identify the specific problem and build the automation that solves it. Not buy software. Not wait. Solve.

How we do it at Studio SmartWork

We don’t sell a tool. We also don’t ask you to learn a new platform. The process is straightforward:

  1. You tell us the problem. We do a quick audit of where time is being lost in your team.
  2. We build the bot or automation. We use proven open-source tools (like n8n) and current AI APIs. We test it thoroughly before putting anything into production.
  3. We leave it running. In less than 7 days, you have automation operating 24/7 in the background. And we take care of ongoing maintenance.

The reason this works —and works fast— is precisely because the science has already been done. We connect the pieces produced by research with the exact processes in your business.

The takeaway

Kornbluth was defending curiosity-driven science because she knows that without it, in 10 years there won’t be new tools to adopt. She’s right to worry.

But the other side of the coin is this: the science of the last 20 years has already produced tools ready to use, right now, in your business. No need to wait. No need to be technical. No need for an R&D department.

What you do need is to identify which process is costing you time and money, and build the right automation.

That’s what we do. And getting started is easier than it sounds: you just have to describe the problem.

Related articles