IN COMPUTER history, 1976 is remembered as the year Steve Jobs and company produced the first commercially available personal computer, the Apple-I, which has evolved into the market trendsetter 38 years later.
But 1976 was also a seminal year in the world of big data and analytics. Another company that started 38 years ago, today leads the world in a rarefied version of business intelligence called predictive analytics.
The main differences are that the SAS Institute is unknown among consumers, and that it has enjoyed revenue growth for every one of those 38 years.
There are great similarities, though, aside from date of birth: both companies were built on a driving vision that remains undiminished decades later; each holds the accolade of being market leader in specific niches – Apple in several hardware categories, SAS in advanced analytics. In parallel with the way Apple put the personal into personal computer, SAS pinned down big data long before it got big.
Today, according to the New York Times, SAS is the world’s biggest privately held software company. And the leadership style of its co-founder and CEO, Dr Jim Goodnight, is the topic of business school case studies.
So, while Dr Goodnight probably wouldn’t appreciate being called the Steve Jobs of analytics, he should take it as a compliment. From the gleaming company headquarters in the unlikely backwoods of Cary, North Carolina, he oversees an operation that continues to set the pace in its field.
Now 71, Goodnight is a man of few words and even less small talk. During an interview at the SAS-hosted Premier Business Leadership Series in Las Vegas last week, he resisted all efforts to be drawn into making visionary statements.
For example, asked about forecasts that emerging markets would overtake the USA as the world’s biggest producer of data in the next few years, he warned: “We’re still not sure if we’ll make use of all that data, but we’ll try. I remember the big RFID thing; this was going to change the world, and all it did was replace barcodes.”
He was referring to Radio Frequency Identification, which allows tiny radio transmitters to be inserted into all packaging, potentially allowing a trolley-load of shopping to be scanned automatically and instantaneously as it is pushed past a pay-point.
Today, it’s become just a subset of the Internet of Things, which revolves around sensors being built into anything that can conceivably be linked to the internet. Goodnight is on top of this trend.
“Pretty soon, everything will have an IP address and you’ll be able to access it from the internet or generate data from it. Today you can turn on your thermostat or burglar alarm via the internet. Cars especially will be using telematics to broadcast all sorts of sensor data every few seconds or minutes.”
It’s not theory, either.
“Right now we’re working with a fleet of 45 000 trucks that will broadcast 60 sensor readings every 15 minutes and then try to make sense of all that data. When you have all the sensor data being thrown out, someone has to figure out what use has to be made of it. We’re in a great position to do that.”
In the developing world, he believes, the pace of data generation will be faster because of the rapid growth in use of cars and appliances, among others.
“In the USA, it will take 10 years to replace current car fleets with cars that have IP addresses. In emerging markets, it’s possible to put in sensors from the start, so we’ll see tremendous growth in the Internet of Things.”
However, he pours cold water on the idea that this will allow for a leapfrog effect, where skipping legacy systems allows emerging markets to move ahead of developed countries.
“Take smart meters in houses that broadcast every 15 minutes what your energy consumption has been. Now we’ve got it, no one can figure what to do with it, except maybe energy forecasting. In future, that meter can send all that data back and inform you to shut down things in the house and tell the air conditioner to cut off or raise the temperature.”
He dismisses suggestions that big data will eventually need some form of built-in self-destruction, due to a rising tide of privacy protests.
“There’s been a movement in the European parliament to have a right of privacy, where no one can keep personal information after two years. That never passed. If you’ve got a 30-year mortgage, what impact would such a law have?”
However, he does not regard himself as a big data crusader against privacy laws.
“Whatever the law requires for destruction of personal information, we’ll certainly comply (with).”
Surprisingly, he does not reject the kind of possibility raised in the movie Minority Report, in which psychics in a “PreCrime” police unit apprehend criminals who intend to commit a crime. Except that the real thing will be based on predictive analytics rather than precognitive powers.
He is asked: “Do you foresee a time when people are taken into protective custody based on the likelihood they will commit a crime?”
He answers without hesitation: “Yes, I do. We will certainly be able to forecast at a high probability level. There is a constabulary in UK that already did that; they were able to forecast that a certain person would be doing a robbery within a certain period after being released from prison, so they followed him and caught him in the act.
“In Iraq and Afghanistan, SAS was at the centre of a programme to forecast the location of IEDs (improvised explosive devices), and was close to 30% accurate. That saved a lot of people’s lives.”
Despite his own ability to see the future, however, Dr Goodnight refuses to be drawn on the possibility of arrest or search warrants being based on predictive analytics.
“It will be up to the courts and legislative system to come up with the rules. I’m not a lawyer; I’m just a developer.”
*Arthur Goldstuck is founder of World Wide Worx and editor-in-chief of Gadget.co.za. Follow him on Twitter on @art2gee and view his YouTube channel at http://bit.ly/GGadgets.