With the value of real-time data access so evident during this pandemic, the conversation on AI has become amplified X-fold across healthcare systems, Research, and Pharma (‘Healthcare’). Great news for many among you. But you can make an even deeper contribution by committing to the drive for ethics.
Ethics relates to how, in the absence of any framework, all stakeholders can work to set clear parameters on the safe adoption and integration of AI – to build towards consistency across real-world settings (clinical and in the community), discovery research, and smart city design.
No-one is suggesting that tech solutions have been sold on hype. But since regulators are deemed to not yet have all of the appropriate skills to evaluate good AI, Healthcare is taking ownership, to scrutinise validity and tech impact across the AI life cycle.
This signals the next chapter in AI’s evolution
We’re now moving ‘beyond the product’ to embed hybrid collaborative systems that draw on human-to-computer interaction. To ensure that through equity, the experiences of the workforce, patients, and citizens are positive.
Regardless of where you’re positioned on the analytics-to ML/AI-to Quantum spectrum, you can offer collaborative support: from grassroots L&D programmes, through to managing knowledge and innovation spillover effects. AI is going local.
AI is Going Local | Some Suggestions for Vendors on how to Support
- Help Healthcare to prep to deploy AI. Those of you that support small data projects across a Tumour Board or multi-disciplinary team can expand your capability into new clinical and operational settings. Can you also support effort on accurate coding mechanisms, retrieving missing data, or conducting data set audits?
- Exploit your own cross-functional team model to tune into clinical workflow more finely, and help identify where AI will create the highest value. Effort is now gearing to focus on co-morbidity, mental health, and social determinants among sub-populations.
- Work to close the perceived gap in language and expectation between clinicians and AI vendors over project management. Time span, team size, and outcomes are often cited.
- Get more involved in hackathons and skunkworks. Healthcare still needs to break through the burden of change from digital-first. So many roles across frontline care have had no training on AI usage capability (not technical grasp), so cannot in turn evangelise to their peers. There is also a clear opportunity for your interfacing sales leaders, and customer success teams to intervene.
- As more cluster and cross-border ecosystem intervention emerges, to exploit multiple EMRs, medtech and IoMT, position DevSecOps as a critical underpin.
- For reference, NHSX in England considers Policing and Criminal Justice as good benchmarks on ethical AI deployment. Policing also shares data with other public sector agencies. Those of you with a footprint can contribute to the expanding NHS Blueprint library.