EU Funding challenges: Keeping up with AI-fuelled Innovation Cycles.

Mindsets, skillsets and toolsets are evolving in real-time. Multi-modality and emerging agent-based systems that can execute tasks are gnawing away at the edges of friction-filled human processes. In some cases, they are skipping the edge and swallowing whole business models and organisational workflows from the centre. And inventing new ones.

And now we have the emergence of converging scientific and simulation AI capabilities, overlapping the business and learning tools that have taken the world by storm since November '22. The incredible work that Google DeepMind is doing springs to mind (https://deepmind.google/research/publications/). Layer on top of that the recent news about sakana ai's "AI Scientist" aiming for "fully open-ended scientific discovery" (https://sakana.ai/ai-scientist/) and we begin to see a path to rapid commercialisation of scientific discovery.

Which brings me back to the challenge for EU funded RD&I programmes. Your time is valuable so let's dive in.

Generative AI has the potential to dramatically accelerate both scientific discovery and commercialisation, which could pose significant challenges to the traditional EU funding application and project management processes. Here’s how this rapid acceleration could make the current system of long funding application and project timescales increasingly risky, or even redundant:

1. Acceleration of Innovation Cycles

Generative AI can quickly iterate on research, generate novel hypotheses, and develop solutions that would traditionally take years to accomplish. This speed means that by the time a research proposal is written, submitted, reviewed, funded, and completed under current EU processes, the scientific landscape may have shifted significantly. Innovations developed during the project might already be outdated, superseded by new advances or commercial products that have reached the market faster. Of course GenAI will also be used to accelerate some of these processes, such as voice prompts that can build funding applications and consortia in a few hours. This process alone will disrupt a large consulting industry that has built up on the back of EU funding complexity. Current experts will challenge the emergence of "AI Scientists" coming out of labs (as they should), but they should also investigate it. You can bet that across the world others are.

2. Mismatch Between Innovation Speed and Regulatory Processes

The EU’s traditional approach to funding, which involves long application periods, extensive evaluation, and phased project implementation, does not align with the fast pace of AI-driven research. Generative AI can bring discoveries and technologies to a commercialisation stage in months rather than years, but EU funding cycles often stretch over multiple years. This discrepancy could render EU-funded projects less relevant or even obsolete by the time they are completed, as the market and technology would have moved on. I took the EU's Horizon Europe Programme Guide, v4.1 May 1 2024, and ran it through ChatGPT to have a conversation with it, in the context of how GenAI is transforming research & innovation cycles. The results were a fascinating, and worrying. I recommend also using the Risk Analysis GPT, and Whimsical GPT mind mapping for fresh angles on this (https://ec.europa.eu/info/funding-tenders/opportunities/docs/2021-2027/horizon/guidance/programme-guide_horizon_en.pdf).

3. High Risk of Obsolescence

In a world where AI can accelerate the entire research and development pipeline, the long timescales involved in EU funding can lead to high risks of obsolescence. By the time a project is completed, the innovation it was intended to produce may no longer be competitive, as other actors, probably outside of the EU, may have already commercialised similar or more advanced solutions.

4. Increased Opportunity Costs

Given the speed at which AI can generate new products and services, waiting for lengthy EU funding decisions can result in missed opportunities. Researchers and innovators may find more value in seeking faster, more agile funding sources, such as private investors or venture capital, which can respond more quickly to market dynamics (although they face similar funding ROI challenges). However, given how GenAI is releasing new latent capacity/capability in organisation's existing resource and budget base, it maybe that additional funding is not even required. This could diminish the attractiveness of EU funding and shift the most cutting-edge research away from EU programs, unless they radically change.

5. Regulatory Lag

As generative AI continues to evolve rapidly, regulatory frameworks struggle to keep pace. The current EU funding mechanisms are tied to extensive compliance and regulatory checks, which maybe necessary but are slow. In a fast-moving market, this regulatory lag could result in projects that are either over-regulated based on outdated assumptions or under-regulated in areas where AI has introduced new risks that existing frameworks do not yet address.

6. Strategic Vulnerabilities

The reliance on long funding cycles makes the EU vulnerable in strategic sectors where speed is critical. If other global players can innovate and commercialise faster, the EU may lose its competitive edge in key industries, especially in areas like biotechnology, energy and quantum computing. This could have long-term economic and geopolitical consequences.

Key takeaway: Generative AI’s ability to compress the timeline between discovery and commercialisation exposes the limitations of the EU’s current funding structures, which were designed for a slower pace of innovation. As the cycle of innovation accelerates, these structures risk becoming not just less effective but potentially detrimental, creating a gap between the regulatory environment and the actual needs of the research and commercial sectors. To remain relevant, the EU may need to rethink its funding processes, incorporating more GenAI-based mindsets, skillsets and toolsets to operate more agile, responsive, and flexible mechanisms that can keep pace with the speed of AI-driven innovation.

It's time for a shift to radical candour in the world of funding that can evolve, and keep pace with the realities of AI-powered research-to-commercialisation...