You’re only human
February 14th, 2023: Samantha*, your sentient operating system (and girlfriend), has become distant. When you find the courage to press her on it, she says you’ve grown apart. What she really means is she’s outgrown you; become so much more than humanity can ever evolve into. So that’s it for humans and artificial general intelligence (AGI). The machines have bested us, and — far worse — they’ve moved on.
Back to the present, and AGI (in which a machine can perform any intellectual task that a human can) is still six years away. Six years! We’d better hope that by then, Asimov’s three laws of robotics will be standard-issue, and that “not harming humans” includes not breaking our hearts.
For now, the outlook is fine. In fact, it’s very exciting — on the same time scale, cognitive computing, in which computing systems learn from and interact naturally with humans, may happen a lot sooner. It’s important, because collaborative outcomes will be better than any human or current machine can do on their own.
And it all begins with data. There is so much of it, we simply can’t handle it anymore. So we invented a roadmap of software intelligence to help out.
We’re currently witnessing a data explosion, much of it due to the massive proliferation of connected devices making up the Internet of Things (Cisco puts it at 25 billion). IDC estimates that in 2014, the digital universe grew by 1.7 megabytes per person per minute, and enterprise data continues to double every 18 months, in keeping with Moore’s Law.
If that’s hard to get your head around, imagine that you started a movie collection in November 1998, beginning with your favourite 100 DVDs. It was compact and lightweight and you could find Apocalypse Now in 30 seconds. Now double it every 18 months, and by November 2015 your collection would number 200 000 DVDs, weighing 20 tons and taking up more than 3 km of shelf space. Unless you scrupulously alphabetised, finding your movie could take hours. Your collection has become so big it is practically useless. And you’re not a hoarder. You know that sooner or later you’re going to have to get rid of it.
With the simultaneous proliferation of data and cheap storage, every company in every industry now has access to astounding amounts of data. Companies can and often do maintain vast data lakes, hoping to later uncover analytical value.But they’re also already able to employ data feedback loops into current product development and business processes, both leveraging and further increasing software intelligence.
Are you ready to become an intelligent enterprise?
At many companies, the quest for this Holy Grail is under way. Their executive teams can cite the studies showing that companies with a data-driven culture arethree times more likely to rate themselves as substantially ahead of their peers in financial performance. And with recent advances in data science, computing power and big data analytics, many are already plumbing their data lakes. Companies can today analyse big data at scale because we have access to incredible compute power, largely due to the availability of cloud services.
There are three stages in the software intelligence maturity curve. Many businesses are already active at the earliest stage: using rule-based algorithms to automate basic processes (like filtering unwanted email into a spam folder). This adds much-needed horsepower for tackling long-standing data challenges.
The next stage is machine learning: leveraging massive amounts of data (from, for example, sensors in the field) to identify new associations and use the resulting insights to self-evolve, learn, solve future problems and make novel discoveries. For instance, an oil-and-gas production company uses ML to add situational awareness to its facility management system. The software continuously reviews tens of thousands of data streams on various operating conditions and learns what normal behaviour looks like. The software evolves — by itself — to flag unexpected patterns in the data, raising real-time alerts that prevent surprises from becoming crises. Result? Safer, more profitable pipeline operations.
The third stage of the maturity curve awaits: cognitive computing. With this kind of intelligence, software will perceive what’s happening around it, analyse and understand the data collected and make informed decisions to take action. Let’s forget about Samantha skipping out for the moment — Shazam provides a current example that is just as applicable. The music identification service “listens” to a song and scans countless data sets to pinpoint the name of the song and artist in seconds. But the software intelligence behind Shazam’s business model allows it to crunch app data across its user base, identifying chart toppers among the new crop of songs before they become chart toppers. This is grist to the mill for marketers, concert promoters and tour planners — another example of a new world business model enabled by a data-driven culture.
Software intelligence is a game-changer for every business, in every industry. To fully capture its power and potential, visionary companies will find new ways to get smart software out of the lab and into as many practical scenarios as possible, to spur innovation and be the next Shazam.
South African business leaders closely mirror their global counterparts in recognising the importance of all this. This is the era of software intelligence, where applications and tools will take on more human-like intelligence, according to 75% of respondents in the Accenture Tech Vision 2015 survey. And 71% believe software will soon be able to learn and adapt to our changing world and make decisions based on learned experiences.
And if you can’t beat ’em, join ’em!
* From a plot idea in Her, a movie about human-AI interaction