It’s an artificial intelligence (AI) special this week. We’re always talking about AI at Gloo HQ—we’re excited to see the potential of this technology as it continues to unfold, and as more and more use cases are explored.

NHS bets big on AI

Last week, the NHS announced it is investing £250m into boosting the role of AI within the service. AI has already made huge waves in healthcare, from helping doctors detect cancer earlier and with higher accuracy to finding new ways to diagnose common conditions.

AI is also helping to improve the overall operational efficiency of hospitals and doctor’s surgeries. And that’s going to make things much easier for healthcare professionals and patients. AI will free up doctors’ time by taking over mundane tasks, allowing them to offer more value to their patients.

As a company that’s incredibly focused on cybersecurity, we couldn’t help but wonder how the NHS is going to deal with this. The increased use of technology is going to bring a whole host of cybersecurity challenges—are NHS staff equipped to handle these? Perhaps cybersecurity will be added as a module to medical degrees in the future.

AI: The Inventor

Researchers at the University of Surrey are seeking to have an AI system, named DABUS, credited as the rightful inventor of two designs. This raised an interesting debate in the Gloo office—should AI be recognised as an inventor or should it be the human that programmed it?

Patent law is complex, but The European Patent Convention does not strictly define who the inventor is. In practice, a human-inventorship criteria must be met in order to receive a patent. This was so that corporations and organisations couldn’t be listed as inventors. But does that mean that AI will miss out too? As we rely on AI more and more, it’s likely it will be inventing future products and systems—is it fair that it doesn’t get the recognition it deserves? It might be time that patent law updated to match the times.

Are you being served?

You’ve been standing at the bar for what feels like an age, watching as the bartender servers everyone but you; you frantically wave your £20 in the air as you try to make eye contact and shout for two beers over the hum of the DJ. We’ve all been there. But experiences like this could soon become a thing of the past.

Pubs in London are trialling new facial recognition technology that will identify the order patrons arrive at the bar. Each customer will be assigned a number and the bar staff will be able to see exactly who to serve next. In true British fashion a queue will form, and we’ll be served on a first-come-first-serve basis. Sounds pretty good, right?

“Artificial artificial intelligence”

Ever heard of “Microworkers”? Not many people have, but recently we’ve seen more and more news stories about this global community. They’re the driving force behind a lot of the AI we’re using in our everyday lives. AI is only as good as the data it has access to. These microworkers are the people getting that data into a format the AI can understand and learn from.

And it turns out that sometimes the microworkers are doing the AI’s job for it. The “Wizard of Oz technique” is where humans are used to perform Human Intelligence Tasks (HITs)—these can be menial tasks like inputting expenses from receipts, or they can be more complicated like identifying the emotions a song makes you feel. In these cases, it’s cheaper to outsource the work to humans than to devote AI resources to it. Jeff Bezos refers to this as “artificial artificial intelligence”.

Usually we see stories about all the jobs AI is going to displace, so it makes a change to hear about the jobs AI is creating. Unfortunately, microworkers are described by many as just another exploited workforce—earning between $0.01 and $1 per task. AI may be creating new jobs, but is it taking all the good ones?

Posted by John on 13 August 2019