The People Business

How can we retain our humanity and continue to use our ingenuity in the age of Artificial Intelligence? That’s one of the core questions under discussion at LocWorld 44 today.

We think it helps when, instead of setting up a false opposition between ourselves and machines, we embrace advancement and recognise the potential of AI to make our lives better.

Hannah Fry has written eloquently of the need for technology to be evaluated for its contribution to society. Today she spoke just as eloquently of the need to recognise the limitations of human decision making.

Hannah’s analysis of judicial decisions was an eye opener. Can we rely on consistency and objectivity from our judges? Not always. Research unearthed some very human flaws.

For example, judges tend to make stricter decisions when they live in a town where the local sports team has recently lost.

How often do people act rationally? How often do we even follow the path of self-interest? When we make decisions we set goals, make plans, gather information and finally make a selection. At least that’s the theory. In reality we take mental shortcuts because we don’t have the time to follow an A to Z routine. Research indicates that the typical adult makes 2,000 decisions per waking hour. Good luck navigating a structured path through that process.

Decision fatigue may lead us to postpone choices or wearily make poor ones. Emotion can dominate our thinking to a point where anger, sadness or excitement may be the crucial and unhelpful reason why we turn left or right at a pivotal moment.

There’s no mathematical model that can tell us which choice will bring us the greatest happiness. As well as gathering current data we need to draw on memory, be influenced by emotion but not overwhelmed by it, and establish context.

And while machines may not be useful channelers of emotion, they can help us draw on memory and they can help us establish context.

Natural Language Processing (NLP) applies algorithms to identify language rules and “teach” these rules to computers. XTM International have taken this process to a new level with the development of Inter-language vector space (ILVS).

ILVS indicates the approximate closeness between source and target words within a language segment. It offers a mathematically precise view of the accuracy of a machine translation, analysing similarities between words across thousands of language pairs and mirroring the linguistic problem-solving of the human brain on a macro-level. Decisions made in the ILVS process are informed by 200 terrabytes of data, harvested from the entire internet. That’s enough information to fill 67 billion single-spaced A4 pages, and the alignments are completed in less than one second. In today’s Q&A Hannah referred to this process, and noted that even the most advanced models fall short of the nuance of human communication.

We couldn’t agree more. ILVS doesn’t replace human ingenuity. It simply offers a calculation that people can rely on, drawing on memory, creating context and giving them a platform to make good choices that use their expertise. It’s a marriage of artificial and human intelligence that acknowledges the limits and showcases the strengths of both. Hannah Fry’s own professional journey has taken her from a world of theory to a far more complicated world in which absolutes don’t apply. In her own words:

“When you’re trained as a mathematician you can admire technology without thinking about how it fits into the world around it.”

This week at LocWorld44 we’re all thinking about how technology fits into and serves the world around it. As Hannah reminded us this morning, and as LocWorld44 will no doubt remind us repeatedly this week, regardless of advances in artificial intelligence, we’re still in the people business. Supporting them, training them, serving them. And business is good.