CIA's Threat Prediction Tech Shows Promise While Raising Questions
It wasn't too long ago that we began exploring machine learning, asking the questions that would propel us into a technological world unknown. Today, in the wake of artificial intelligence, one resurges with relative frequency: When will this technology go too far? The CIAs recent revelation resurrected that concern.
Mainstream Machine Learning
Machine learning algorithms as computational concepts are old news, but their capabilities as tools are ever-evolving. The relentless efficiency of machines to learn from repeatable processes, faster and faster, is a game changer. We see it deployed on social media and online markets to great effect.
For example, if you have liked a particular holiday destination, brand, or TV show but avoid politics, predictive analytics will plot the patterns to give you exactly what you want. This is a double-edged sword. For one, it creates a veritable echo chamber, limiting our exposure, and reinforcing our preferences. It also limits our worldview in an ecosystem predicated on expansion.
Albeit important, the social phenomena are today overshadowed by news that the Central Intelligence Agency (CIA) harnessed deep and machine learning techniques to create new forms of information gathering and possibly pre-empt security threats.
The CIA’s Crystal Ball
The CIA developed tentative intelligence known as “anticipatory intelligence,” capable of predicting social unrest five days before it happens.
By gathering billions of bits from disparate sources, the CIA runs sophisticated algorithms and analytics to identify patterns correlated with potential threats, civil unrest, protests and riots. Whether key targets are mostly civilian or governmental is unclear; and whether we’ll feel protected or invaded is to be seen. Whatever the case, some will certainly be discomforted with good reason.
The news is reminiscent of Sci-Fi — a recurrent sentiment with each passing day. Minority Report explored the concept of anticipatory (clairvoyant) crime fighting and illuminated the pitfalls, such as undue confidence in a system of possibilities, wherein real people suffer for imagined crimes.
Selling Dreams
The tech industry as a whole is imaginal, selling us solutions to problems we didn’t realize we had. It’s a self-fulfilling prophecy and we are along for the ride, unquestioning and purchasing. The recent wave of smart-home devices is a case in principle. I’m sure you hadn’t realized just how much you needed a voice-controlled smart-speaker that could regulate your lights and share arbitrary facts with you about aquatic animals until now.
This kind of integrated technology is a natural consequence of collecting and analyzing big data from everyday people. Tech companies know so much about us that they believe they can anticipate our needs. Now the CIA will anticipate threats to our security.
Well, if these situations are analogous, the CIA may similarly struggle to unlock tangible results and extract real value out of the intense white noise. Yet this is an ever-evolving technology. It’ll get better.
Tech Oracles for Business Success
Imagine for a moment that you knew what your customers wanted to buy before they did. Without the budget to compile and mine big data, how would your competitors compete? That’s the current business reality. And it isn’t just the competition that’s changed; it’s the consumer-base.
Consumers expect everything on demand and tailor-made. There is a sense of urgency, a demand for immediacy, and an expectation of personalization. These axioms will determine the next big changes for society, business and now government.
In Sum
Electric cars are mapping and learning from the roads that they traverse and companies know exactly what users are saying about them online. Smart-home devices are now in our homes, listening, collecting, analyzing. The CIA is now watching out for possible security threats.
Undeniably we are scratching the inexorable surface of a social revolution that will ripple through the entire world.
These are the questions that businesses and people need to consider moving forward.
- Nominate me for LinkedIn Top Voices List
- Follow me on Twitter
- Other Articles by me on the Topics of Leadership | Self-Improvement | Tech | Digital | Careers | Entrepreneurship | Innovation | Sales | Marketing | HuffPost Column | Forbes Column | General
Li-ion Cell Manufacturing Process Operations & Technology. Physicist.
8 年Albert-László Barabási - Bursts; Indeed read of any of this guy's books. Our behaviour patterns are on the whole predictable, whereas the random outliers need further analyses.
Disrupting the Status Quo with Agile | Human-Centric Change and Learning Experience Design
8 年This is a really simple one here. Obsession with the wizardry of data and technique, as Edwin Friedman insists, often blinds not illuminates. It becomes a form of addiction, which turns professionals into data-junkies, and their information into junk yards. Here ends the lesson you all.
Homo sum humani a me nihil alienum puto.
8 年There is this belief that, if you just have enough data points you can predict large scale events near term. That may not be so. If anything, there may be a few key indicators that predict the magnitude of what is going to happen - within a rather big time window. The CIA has been periodically criticized for failing to predict big events; I can appreciate that they are trying harder, but the difficulty isn't always appreciated.
Realising real value from AI and data investment | SF writer
8 年Actually, the American TV series Person of Interest explored this scenario, albeit in a sensationalized form. One of the key themes that came out of that show, and is being born out by some of Microsoft's ill-fated AI products, is that human's set the framework for these machine learning tools. Even if they evolve independent of humans, the initial framework they start from bakes in a very human bias. As a result, I suspect the predictions that the CIA receive from their AI will conform to their world view.