Zephyrnet Logo

We gathered together some Register-reading IT pros to talk about voice AI assistants in the legal, finance worlds. Here’s what they said…

Date:

Sponsored It seems a fair guess that this year we will see more voice-enabled goodies under Christmas trees than ever before. And while a generation of youngsters will, thus, grow up with this voice-enabled tech, they are largely heading towards a workplace where the keyboard remains king when it comes to communicating with fellow workers.

So, with that in mind, it seemed entirely appropriate to bring together a bunch of Register readers in senior IT positions, in a very Christmassy venue in the UK capital last month, to enjoy a slap-up breakfast and discuss how voice is making itself heard in the workplace.

The venue was atop a skyscraper in the City of London, and unsurprisingly, the group was dominated by execs from the finance and legal sectors, either end users or suppliers running services on their behalf. All were interested in voice-controlled technology, and were at least familiar with the use of voice-driven applications in customer-facing situations, in call centres, for example. But it seemed few organisations had implemented voice-recognition software internally as an alternative to the keyboard for generating reports or composing emails.

One attendee said financial organisations might use voice and AI to “reduce the first level of interaction,” serving up, for example, a series of questions to identify customers with a relatively small amount to invest and directing them to a particular portfolio, while those with larger sums – eventually – find themselves talking to a human adviser.

…they want to take human beings out of the process flow

“I’ve seen my government clients starting to introduce it into the front end of the contact centre,” said one procurement specialist. “Really to save money because they want to take human beings out of the process flow.” Despite this, he added, when it comes to implementing newer technology, “government is quite conservative.”

And that was a recurring theme throughout the conversation: conservative – and that’s small-c conservative – leaders in organisations who don’t see the need to bring voice inside the enterprise, or indeed any new technology.

The legal sector seemed to be a particular offender in this regard. One attendee described how university students were used to using digital recordings of lectures, even searching for keywords: “It’s completely part of their daily makeup,” the exec said.

Those same bright young things then end up in law firms, where “you’ve got the 75-year-old partner who says, ‘I’ve always done it this way, so this is the way we’re always going to do it.’ And they just hit that brick wall, and you see these poor associates printing stuff out and writing on them and giving it to a PA to type up. It’s that cycle that I’d love to break.”

Another attendee with previous experience in legal and financial firms was agog that this was still going on, with the first explaining that, yes, there are still “banks of legal secretaries,” including those who do night shifts to pickup the slack, all waiting on lawyers to dictate letters and other communications, “line breaks and all.”

Yes, those senior partners will eventually move on, our frustrated exec continued, but so does technology, meaning the legal sector in particular will forever lag behind when it comes to adopting tech that would streamline the workplace.

Even simply reducing the amount of printed matter within an organisation could improve efficiency, one attendee said, just in reducing the billable minutes sucked up fixing clogged printers.

There are other generational differences that may have a bearing on whether, and how, organisations adopt voice-powered technology. As our procurement consultant said: “My children don’t see a problem giving their personal data away, and they’re in their thirties.”

Another pointed to other trends among younger employees: “We’ve got apprentices in our workplace, 18- or 19-year-olds born this century, who don’t use phones. It’s kind of baffling. To us it’s basic, but they’ll send an email, rather than picking up the phone.” This all suggests that using voice-to-text will not be a leap for younger employees.

However, concerns remain. One attendee said many people are wary of interacting with voice systems in everyday life. If your law partner is wary of using voice-controlled systems to book a GP’s appointment, are they going to be more open to the technology, just because it’s provided, and even made mandatory, in the workplace. “We can change it, but we have to get people to accept it,” they said.

More broadly, the group were keen to discuss privacy and security, both as a factor in how employees will react to using voice, and where and when they will use it, and how companies can ensure their systems and data are secure when voice is implemented.

This included issues such as where the data is processed, and whether it is moved offsite, or even off-shore, and what the implications were for GDPR obligations.

Air gaps

One attendee questioned whether the need to update training data for voice-powered systems would make it difficult to have completely air-gapped systems, raising privacy and security concerns.

And one finance sector tech pro questioned whether processing voice will raise additional compliance issues. Someone working in a regulated industry might have 50 or more compliance regimes to take account of, they said. Things could become particularly difficult if an organisation is recording conversations where another party is involved.

“So, you may decide to use it. But just by doing that, you might find the costs outweigh the benefit,” they speculated.

And while email text is measured in kilobytes, another attendee said, non-trivial voice recordings will be much larger, raising a content management issue.

A representative of Nuance, which sponsored the breakfast get-together, chipped in here, saying that its Dragon service is cloud-based, with its cloud provider taking care of GDPR and security issues, while the underlying algorithms are regularly updated in the cloud, meaning no overhead for customers.

All of these questions lead into perhaps the most important issue: as with any technology introduction, how do you ensure the project is successful and actually adds value by making your employees more productive?

As one attendee queried: “Does your voice solution solve a problem? Is it labour saving, does it add productivity? Does it give time back to someone?” And as our finance pro adds, do the benefits outweigh any additional overhead?

One attendee, tasked with exploring new technologies for their outfit, agreed, saying organisations needed to be clear what they intended to use voice for: “Is it for an immediate ‘I want something now and I’m going to say it’ in which case, do you really need to record it? Or are you using it for security purposes, in which case, it’s part of your toolkit but not the only thing you can rely.”

Unfortunately, this is not always the case. As one veteran said: “I see many different IT operations deploying things that are haphazard, that do not relate to any business strategy, that do not seem to be based on a technology strategy, really. And that plays out in that your technologies don’t work on the desktop. That all relates back to the thinking on how we run IT.”

As for Nuance, its representative was quite clear that a successful voice strategy depends on strong leadership, both in terms of overall business strategy, IT strategy, and voice strategy.

Others focused on how adopting voice processing in enterprise workflows could change the nature of what was being said. “Am I better writing it down because I’m able to structure my thoughts or if I just say something into a dictation device,” pondered one attendee. “The way you say something against the way the you would write something, there’s an absolute difference.”

So as we came to the end of the session, had our group changed its mind about the potential of voice in the enterprise?

‘Ticking time-bomb’

One developer said that having used voice systems in the past, “hopefully, we’ll be more optimistic about maybe trying it again.” More specifically, they saw it as a viable solution for “for something I see as a ticking time-bomb” in terms of the physical impact of constantly using a keyboard. “I work in offices where people are hammering away all the time. And eventually people get sore necks, sore hands. And sometimes you don’t catch that early enough.”

One exec floated the idea of voice being able to speed up specific workflows, citing the possibility of a finance exec seeing a bank in trouble and being able to navigate to the relevant data in their systems to determine their potential exposure, “almost in real time instead of sending a whole team scrubbing around, trying to drag that information out of different systems.”

Another, from the legal sector, envisaged voice making it easier to deliver help and support to users: “With the services I manage, we struggle to find talented people who actually care, and who know where to get the information.” How much easier would it be, he continued, if voice and AI could combine to deliver the right responses to problems that have been given over the phone, or even by email.

That voice technology would improve leaps and bounds seemed to be generally accepted. One attendee recalled a recent conference where a group of presenters were talking in multiple languages, with almost instantaneous transcriptions and translations being delivered on screen. The implications for organisations running multi-national teams are obvious.

That said, they continued, you can have too much of a good thing. “I don’t want the Sky or BBC subtitles to change – because they’re hilarious,” said one.

Breakfast roundtable sponsored by Nuance.

spot_img

Latest Intelligence

spot_img