Over the Thanksgiving weekend, I took my daughter to visit the Computer History Museum, just around the corner from the Googleplex in Mountain View. It's a terrific museum that gives real context to the massive change we've seen in technology and our everyday lives over the past 50 years.

One of the highlights of the museum is the ability to learn about some of the 'forgotten' heroes of the computer age. One of these was Seymour Cray, the inventor of the CRAY-1, which between 1976 and 1982 was the fastest supercomputer in the world. It looked like the future then and, in many ways, still does now. Seymour was the archetype of the 'wacky inventor' who had a passion, among other things, for building underground tunnels in his home and had some ingenious ideas inspired by the most unusual things for how to make computers work faster.

But what struck the biggest chord with me was learning about the three simple principles he stuck to when building his company culture and creating teams to make new breakthroughs in computing: start fresh with each new project; design simply using proven technologies; and work in small groups with a single decision-maker. These seem deceptively simple; obvious even. But I think they struck a chord because they feel so alien to how we, as an industry, have a tendency to work. Let's take a look at each of them.

First, start fresh with each new project: Cray had a simple belief that you were not going to find new solutions by approaching a different problem with the same approach, mindset and process every time. He was an enemy of repeatability (read predictability), a huge proponent for curiosity and the liberating potential of a blank sheet of paper.

Compare this with how agencies normally work. We look for inspiration in the same place – other ads (look at what is mentioned in meetings and the dreaded competitive review); we use the same briefing formats; and we sell process (aka how we make the sausage). And we do this regardless of the problem we are addressing. We don't like talking about doing things differently – it sounds dangerous and uncertain – and we will rarely admit the simple truth that different ideas require different people working together in different ways. Perhaps we need to apply some creativity to the way we work in order to make our work more creative. Maybe we should spend time at the start of a project thinking about how we can approach the problem in a fresh, new way – one tailored to the task at hand.

Second, design simply using proven technologies. There's two ideas in this principle. The first is the notion of Occam's razor, namely that the simplest answer is the best answer. We have a tendency, despite years of being trained to 'Keep It Simple Stupid', to layer complexity on top of our ideas in a vain attempt to make it look new. Perhaps to create something new, we need to strip away everything superfluous from the idea. (The same is true of the ways we work.) The second notion is that of 'using proven technologies'. This made me think of our ambivalence still to build on things that already exist. We try to build the new from scratch far too often and rely far too little on technologies and platforms that already exist. The imaginative and interesting use of APIs, for example, is still a relatively underdeveloped field of creativity in our industry. Maybe we need to stand on the shoulders of others a little more often.

And third, work in small groups with a single decision-maker. The final principle runs absolutely counter to how we work. Our remuneration systems, more often than not,encourage us to work in larger groups so we can charge for more (preferably senior) FTEs. The work this large group eventually produces is then run through layers of hierarchy in the agency and then again through the committees and layers of the client organisation. It's amazing anything makes it into the world and not surprising that so much that does appear has no edges left.

We all know the best work comes from a small team with an empowered decision-maker – it's made the best ads and is how the tech industry tends to run (there's an Amazon rule that no team should be larger than it takes to be fed by two pizzas). Better still, build small teams with real diversity of experience and expertise and there's a far greater chance of the magic of the new happening. We would all do well to remember David Ogilvy's brilliant aphorism: 'Search your parks in all your cities, you'll find no statues of committees.'