Springboards launches Flint to tackle AI’s sameness problem

Artificial intelligence is getting smarter, faster and more polished. But in the process, it may also be getting more predictable.

That’s the problem Springboards is targeting with the launch of Flint, a new AI model designed to generate more varied, less obvious creative outputs, as concerns grow that large language models are converging on the same answers.

Springboards, which has built a following among more than 100 agencies globally across the US, UK and Australia, is also opening up its platform beyond its traditional subscription model, making it available to a broader base of creatives, freelancers and marketers.

“We don’t believe that great creativity will ever come from a machine,” Pip Bingemann, co-founder and CEO of Springboards, said. “That will always be a human thing.”

SEE MORE: How Noosa startup Springboards is reshaping creativity

Instead, Bingemann said the company has focused on building tools that help people think more widely, rather than automate the creative process.

“We’ve been working on this in secret for well over a year, to work out how you train a model not to give you the same answer as everyone else,” he said. “The goal is to go deeper, to make it more inspiring and get different ideas out of a language model.”

The problem with “smart” AI

Flint sits at the centre of that approach. Described as a “divergence model”, it is designed to prioritise variation over consistency, a deliberate departure from traditional large language models, which tend to converge on a narrow set of familiar responses.

“For a lawyer or an accountant, convergence can be a feature,” Bingemann said. “But for a strategist, writer, marketer or creative team, it’s a bug.”

That convergence, he said, is becoming more pronounced as models are optimised for accuracy and reliability, often at the expense of originality.

“Many teams are stuck between two bad options: a slow strategy that holds everything up, or skipping strategy entirely and going straight to production,” Bingemann said.

“In a world where creative production is fast and cheap, deciding what to make matters more than ever.

“The goal isn’t slower or faster, it’s always-on: strategy that’s embedded in the workflow rather than a separate phase.”

Training for variation, not sameness

Flint has been built on a lightweight, open-source foundation and favours speed and iteration over scale, with the aim of generating a wider spread of ideas at the earliest stages of thinking.

Rather than refining toward a single “best” answer, the model is designed to explore multiple directions, surfacing less obvious creative territory.

In testing, Springboards said Flint significantly outperformed leading models on creative diversity, producing more distinct outputs across repeated prompts.

Kieran Browne, chief technology officer and co-founder, said the focus on novelty reflects a shift in how AI needs to be applied in creative industries.

“The reality is that frontier models are prioritising accuracy and correctness over originality,” he said.

“Flint is built on the belief that human taste and creativity must be at the core of good creative work; we are optimising for variation rather than automation.”

Opening up beyond agencies

The launch also marks a broader evolution for Springboards, which was founded by Bingemann, Amy Tucker and Browne and has spent the past three years building tools for agency and in-house teams.

The company operates across Sydney, New York and London, with around 20 staff based locally, and has historically focused on agency clients. That is now changing.

Springboards has introduced a redesigned platform and new self-serve plans, including free and paid tiers, opening access to freelancers, smaller teams and in-house marketers.

The platform itself has also been rebuilt, shifting toward a more flexible, collaborative interface that allows users to move from early-stage insights through to briefs, concepts and execution within a single workflow.

Tucker said since day one, customers have been at the centre of its innovation.

“Our goal has always been to build tools that enable advertisers and marketers to do their best work, and this new platform is the culmination of that,” she said.

“We’re so excited to finally open this up to everyone, from solo freelancers to global agency teams. Whether you’re a strategist, a creative or marketer, you can now use our platform and model to explore your best ideas with expert support at every stage of the creative process.”

Creativity still belongs to humans

Alongside the product shift, Springboards is also investing in training, launching a monthly masterclass series featuring industry figures including Mark Pollard, Faris Yakob and Zoe Scaman.

The aim is to pair AI capability with human judgment, rather than replace it.

“We optimise for variation to help people go really, really wide, and then let them decide what’s good,” Bingemann said.

That philosophy sits at the core of the company’s approach and its view of where AI is heading.

As content becomes faster and cheaper to produce, the real challenge for marketers is no longer output, but originality. And in that environment, the advantage may not come from better machines, but from better thinking.