NHS Alexa Team Up

It was interesting to see the reaction to the recent team up between Amazon’s Alexa and the NHS. In principle the idea is pretty straightforward: ask Alexa about a symptom and Alexa – that clever little voice that comes out of Amazon’s Echo devices – tells you what could be wrong.

Technology like this is a great example of how people can access reliable, world-leading NHS advice from the comfort of their home, reducing the pressure on our hardworking GPs and pharmacists.
– Matt Hancock, Health Secretary

 

Perhaps you want to ask about what to do when you get a chalazion, or how to bring down a fever. Alexa now has all (well… most) of the answers thanks for a collaboration between Amazon and the NHS. Originally, the NHS has been very keen to get online advice content accessible and correct. A huge amount of effort by healthcare professionals has gone into creating online content such as their NHS Conditions site, which continues to be an incredible resource for the general public in the UK and beyond.

Amazon’s AI algorithms have effectively crawled and “understood” (we use that term very lightly) the NHS website content so that it can now be verbalised for the elderly and hard-of-sight who among the least likely to use a web browser let alone look up medical conditions. Now with the Echo Dot, example, they are able to simply ask Alexa verbally.

Is there really anything clever here?

This is a difficult question to answer. On first glance it doesn’t look too far from verbalising web content. Heck, we could do that in 1989 with an Amiga 500.

Certainly both Amazon and Google – and of course Apple with Siri – have made great strides in digitising conversation-based systems through their various gadgets. However, it is not without its problems. There have been widespread reports of bias in training, such as Alexa finding it difficult to decipher heavy accents. That aside, there is also the complexity in the ways we all speak. We often insert slang into dialog, ask the same question in dozens of different ways with different intonations, and you’re probably familiar with that knot-in-throat scenario where you suddenly become very conscious of the question you’re asking of a machine, and your words become so mangled even a human couldn’t tell what you’re trying to say! These are the things an “AI” needs to deal with.

And then of course there is also a problem of language itself. More than 4.2 million people in the UK spoke a foreign language as their first tongue in 2011. Almost 1 million of those do not speak English well. The number is likely to be even more now. It is not clear whether the service will extend to that corner of the population.

Sound great…?

On first thought, yes – but actually there has been a huge reaction to this announcement. Aside from the cynical criticisms, there has been a backlash regarding data privacy. This can’t have been unanticipated. There has already been a deep grumbling on social media about Amazon’s ability to continuously listen to what’s going on, and with Cambridge Analytica news hitting the headlines earlier in the year, and in light of solidification of GDPR last year, the general public are now far more aware of their right to privacy and potential for privacy leaks.

CNBC

An outsider’s perspective of Alexa’s dip into digital health

Digital Health

The downside of digital health: Privacy concerns and risks

In our opinion, the concern is a valid one. At the time of the announcement privacy hardly had a mention so it was natural that this would be raised. However, since the initial announcement @NHSX publicised some more information, which highlighted – importantly – the terms of the data deal. Data will not be passed on or sold to third parties. That is a massive step in the right direction, so long as there are no caveats! Whilst privacy concerns are of course vital, we think NHSX has done a good job in normalising a conversation that could easily descend into farcical fear.

Amazon has been very clear that it is not selling products or making product recommendations based on this health information, nor is it building a health profile on customers. All information is treated with high confidentiality. Amazon restrict access through multi-factor authentication, services are all encrypted, and regular audits run on their control environment to protect it
Eva Lake, NHS Digital

Indeed, the privacy argument was nowhere near as vociferous before the initial announcement; when you type your embarrassing condition into nhs.uk, it is highly likely Google and/or some other data giant is collecting your browsing history. The difference is of course you can always clear your data history in those systems and at least be semi-confident that that data is now gone forever. We accept there is an element of chaotic ripple effect that could never truly be deleted, but at least that is something.

The assurance that data will remain private and not salable or attributable to a particular individual goes a long way in allying the fears the general public have when you put the words “AI” and “health” together. There needs to be a coming together from the general public on one side requiring a re-education on what AI really represents and technology companies on the other side taking very serious and committed steps in assuring privacy.

Manish Patel
manish.patel@jiva.ai

CEO @ Jiva. Stringer of numbers in complex patterns.



Latest News

EU AI Act Developed by the Organisation for Economic Co-operation and Development (OECD), the EU AI Act is the first legislative proposal of its kind in the world and supports a global consensus around the types of systems that are intended to be regulated as......

Singapore, London, 06th February 2024 —   Jiva.ai, a leading no-code AI platform, and Aevice Health, a prominent provider of remote respiratory monitoring solutions for the healthcare continuum, today announced their collaboration on a jointly funded co-innovation programme by Innovate UK and Enterprise Singapore aimed at......

Looking to 2024’s innovations The age of language models is here (sorry to state the obvious). 2023 saw the advent of easy-to-use LLMs, and the rise of their pre-existing underlying enabling technologies – transformers and retrieval in particular – which transformed the AI landscape in......

Social