EU R&D&I Funding...GenAI Signals to Watch Out For...

That's Research, Development and Innovation. Your time is valuable so lets dive in...

1. Time and cost compression will (needs to) play a big part in all aspects of EU funding programmes. This will be a result of reducing friction in the system. From the bartering and allocation of funds to programmes, the design of "calls", the funding proposals put forward by consortia/individual applicants, the evaluation process, the negotiation process for receiving funding, through to the actual R&D&I work itself (including exploitation of results). There are evolving, compounding percentage savings everywhere, and these will rapidly fan out across the various parties. Baselines for project costs will continue to be reset along with critical paths. Old assumptions will be cast aside, and emergent properties will bubble up across all aspects of the funding lifecycle.
2. The AI doesn't do stuff on its own (not yet anyway - see Agents). Rapid performance improvements will be led by artisans, both within the existing broad ecosystem of EU R&D&I funding participants, and from new disruptors, able to quickly and cheaply exploit opportunities to remove friction in the system.
3. Funding will go further. Much of the effort in the funded projects goes towards manual activities such as meetings, discussing, ideating, planning, budgeting, organising, travelling, various forms of iteration etc. All of these are ripe for friction reduction, and this will happen from within the projects, driven initially by participating partners discovering more and more possibilities latent in LLMs, not to mention the integration of more AI capabilities into standard productivity tools. Efficiency savings should not necessarily be translated into budget cuts. The savings can be redirected towards more of the science and innovation elements of the projects. For more thoughts on GenAI latent capabilities read Ethan Mollick's blog, Latent expertise: Everyone is in R&D (https://www.oneusefulthing.org/p/latent-expertise-everyone-is-in-r)
4. Voice, video and simulation capabilities will become key components of funding proposal construction, evaluation, negotiation and project implementation. This will positively impact both the scientific, exploitation and bureaucratic aspects.
5. Research Institutions/Universities, Corporates, SMEs and Startups will be the key beneficiaries of the acceleration of GenAI penetration into EU-funded programmes. And so they should be. Costs and risks of participation will come down, and by taking advantage of emerging simulation techniques, scientific and business model discussions will become more engaging and less risk averse. Yes/No participation decisions will become faster and less resource intensive (cheaper).
6. Navigating the EU R&D&I environment is a complex evolving process and it has given rise to an entire industry of consulting organisations and advisors. These converge at the EU funding programme level (deciding on calls, technology trajectories etc.), preparation of applications and consortia building, negotiation of contracts, and coordination of funded projects. There is friction everywhere, and that means opportunity. Time, cost and quality performance improvements powered by consultancies harnessing generative AI will continue to put pricing pressure on these business models. However, these players also sit on a lot of valuable data, insight, contracts and business & technology relationships across vast numbers of public and private organisations around the EU and associated countries. Those consultancies with deep technological/scientific resources (humans) and data have opportunities to pivot to new services before pricing of current offerings becomes a race to the bottom.
7. Funded projects already underway will continue to exploit evolving generative AI capabilities, and this will allow them to potentially achieve their planned milestones faster, with higher quality, and at a lower cost. How this will exactly play out is hard to predict, given the nature of contracts and bureaucracies involved. But more of the budget should be made available for the scientific and exploitation aspects of funded projects vs the reporting and coordination which can/will be automated away to a large extent. Because new calls are coming out all the time, and overlapping with funded projects in progress, this will be an interesting puzzle to solve with many feedback loops.
8. Time and cost to commercialisation of funded projects will come down, indeed it will have to. Given the overlapping trajectories of many of the underlying technologies that are funded, speed is a key. Many EU calls allocate funds to three year projects (+/-), and on top you need to account for the time it takes to propose, get a positive decision, and then negotiate across large consortia before the project can actually start. Generative AI, robotics, bio-tech, material & molecular design, energy, mixed reality and many other foundational technologies won't wait for those timescales. Expect to see some rethinking on this both at the EU level, and also within organisations considering participation in funding projects with long timescales.
9. Generative AI (and eventually Agents) will be used to help generate funding calls, evaluate proposals, negotiate and track funded projects. A recent article in Science/Business, a commentary on EU industry, science & policy, mentioned the forthcoming new initiative, GENAI4EU. It will be interesting to see how GenAI is used to facilitate this, as the technology continues along its own exponential curve/adoption rate (https://sciencebusiness.net/news/r-d-funding/commission-shaping-eu500m-generative-artificial-intelligence-calls).
10. Unsuccessful funding applications. Much effort (time/cost) is spent on applications by consortia, individual organisations and consultancies that do not get funded. Yet many of these applications are of high quality and develop powerful partnerships and possibilities during the proposal building phase. Sometimes these unfunded teams resubmit the proposal in the next funding round, but that has a big time constraint, and meanwhile the technology clock is ticking. Watch out for opportunities to harness unfunded projects, perhaps through new forms of highly automated brokerages, and different approaches to financing.
11. Next EU Framework Programme (FP10). Work is underway on the programme that will succeed Horizon Europe (https://sciencebusiness.net/news/fp10-should-boost-industrial-competitiveness-european-associations-argue). Again, technologies are racing ahead at such a rate that the EU will need to utilise generative AI and simulation capabilities to ensure that it can adapt to the realities of the world in 2027 when it kicks off. The European Association of Innovation Consultants (EAIC), a lobby group, has created position paper with 8 key points, many of which contain a lot of opportunity for removing friction by applying evolving generative AI/Agent technologies. You can download the EAIC report here and upload it into ChatGPT for analysis (https://www.eaic.eu/news/eaic-position-paper-on-fp10).
12. Global competition: expect the US, China, UAE, KSA, India, African nations, South-East Asian nations and Latin America, to continue to power ahead with how they utilise evolving forms of generative AI to accelerate R&D&I, attract talent and drive economic progress.

Key takeaways:

• In 2 years the pace of EU R&D&I funded projects will be unrecognisable from what it is today. It will have to be.
• Evolving forms of generative, and other AI's, including Agents, will massively reduce friction and cost in the system, disrupting incumbent organisations and opening up new markets for others.
• More focus and a higher percentage of the funding will gravitate towards the actual scientific and innovation exploitation elements of the programmes as automation takes its toll on bureaucracy.
• Universities, national Research Institutions, Corporates, SMEs and Startups will be the key beneficiaries.
• Evolving mindsets, skillsets & toolsets are critical to keep up.