Beyond AI: The future of Intelligence is collective (human and machine)
What would you describe as intelligence; memorising notes for exams only to forget about them in the summer, or making funny jokes on the fly on seemingly any topic?
How about a magician who performs strictly on a script or a magician who changes the routine ever so slightly depending on the audience?
This isn’t a case for who is ‘more intelligent’, but rather, when something is an act of intelligence.
Intelligence is a widely used term (“So and so is highly intelligent”, “emotional intelligence”, “artificial intelligence” etc.) across many aspects of life. It’s often associated with the ability to learn and process information quickly, broadly or deeply. But is that all there is to intelligence? We all have some idea what intelligence looks like, but yet, we struggle to explain clearly what intelligence is.
In this blog I explore ‘intelligence’: why it is critical for decision making, especially considering the limitations of the human mind. I also look at (the lack of) intelligence in the context of the information-overloaded world today and how it’s heavily misused or misrepresented in the market. Finally, I share how Human Managed creates intelligence for our users and what we look forward to about the future of intelligence.
What is intelligence?
If you thought making funny jokes on the fly and the spontaneous magician is intelligence — I agree with you.
But why? Is intelligence the ability to improvise?
Not quite, but effective improvisation is only possible with a good grasp of relevant knowledge. It’s a result of having a depth and breadth of knowledge and knowing when to apply the right knowledge. It’s spontaneously creating in the moment by using, combining or iterating prior knowledge about something.
This is not the same as someone just ‘winging it’ without understanding the topic, or someone just making things up from thin air.
So…
- Intelligence is applied knowledge. Like applying what you learned in school to solve a problem at work.
- Knowledge is stored and accessible information or learnings. Like what you remember from school.
- Information is contextualised data. Like the revision notes that you created.
- Data is units of information. Like the resources you read during research.
When information is processed, analysed, knowledge is created. And only when a relevant knowledge gets applied to improve something or achieve an outcome, that is intelligence.
Decision making with intelligence
Decision making is the process of making choices from available information or intelligence, and good decision making can be aided by accurate and useful intelligence.
We use output from human and machine analysis to make decisions.
Human Intelligence
Human intelligence — or applying the right knowledge at the right time — is no easy feat. You can master a skill or a topic but when it comes to decision making, there are so many variables at play — starting with your own mind.
And it’s probably no news that the human thinking process is not foolproof — despite its miraculous brilliance, it also has inherent weaknesses and biases.
- Economist, political scientist and cognitive psychologist Herbert Simon coined the term “bounded rationality” in the mid-1950s to describe human rationality as bounded by limitations of time, data and processing power.
- We operate with simplified mental models of reality, because the mind cannot cope directly with the complexities of the world. We use them all the time — both intuitively and purposely (for example, second order thinking: “If I did x, what will happen after that?”).
- Psychologist and economist Daniel Kahneman divides the human brain into two figurative agents, System 1 (fast, intuitive thinking) and System 2 (slow, deliberate thinking). Our brains are highly evolved to perform many tasks on ‘autopilot’, and are only sometimes activated to carry out logical, analytical tasks. In both Systems 1 and 2, thinking is impacted by behavioural fallacies (eg. availability bias, law of small numbers, social proof, framing, anchoring)
And the human mind’s limitations do not just affect me and you. Richards Heuer (a CIA veteran) argues that everyone — even experts like CIA intelligence analysts who are highly trained in rational thinking and decision analysis fall victim to the same analytical traps. In fact, he says when it’s the experts that fall victim, the effect can be much worse because they attach more confidence to their own expertise (their self-perception or from others).
So yes — human intelligence is hard, because our minds perceive data and process information with our own biases and mental models.
Machine Intelligence
Machine intelligence is simulation of human intelligence in machines that are programmed to think like humans.
The role that machines play in human decision making is becoming more commonplace due to the advancements in:
- artificial intelligence (techniques that enable computers to mimic human intelligence like logic, if-then rules, decision tree, etc);
- machine learning (where machine learns from data and builds predictive models on the fly without programming); and
- deep learning (machine learning with functions that operate in a nonlinear decision-making process from structured or unstructured data, with or without supervision).
Your divided attention, please?
Ok, so what? Humans have always managed to progress, even with our innate limitations, right? And now that we have machine intelligence that’s great at repeated analysis, decision making should be easier, right?
You guessed it — Nope.
Decision making is not only influenced by human’s cognitive capabilities (which then influences machine’s learning capabilities), but also the situational context and environment. What is going on around us (physically, digitally, socially) are all important determinants of our behaviour.
In 1971, Herbert Simon spoke about “information overload” that comes with technological advancement such as magnetic tapes, remote consoles wired to computers and large data banks. This wealth of information created what he coined an “attention economy” — where the human attention is a limited, scarce resource.
“A wealth of information creates a poverty of attention” — Herbert A. Simon
Today, half a century later, the attention economy is the rule of the game and seemingly unstoppable, with tools, applications and algorithms competing for a slice of consumers’ attention from every channel.
The ‘So What’ is this: today’s data-and-information-flooded and attention-hungry environment, combined with the inherent weaknesses in human thinking process is not optimal for well-grounded decision making.
Are you really getting intelligence?
Previously, the general focus on digital transformation seemed to be that there was not enough data to be “data-driven”, or it was too costly or technically difficult to collect and explore data. With the democratisation of cloud computing and off-the-shelf tools that collect data, the problem is then not so much the input of data, but the processing of information to produce meaningful output.
Here is the irony: today, the market is filled with solution providers are claiming to provide better and faster insights or intelligence for decision makers, but some organisations do little or nothing with data to aid their decision-making.
There is an underlying and dangerous assumption that if there is artificial intelligence involved, decision making will get easier.
This is simply not true, for many reasons but especially because:
- Machine ‘intelligence’ that does not enrich data with relevant context is merely presenting information, leaving humans to interpret whether something is relevant knowledge (death by dashboards, anyone?).
- Most solutions are built for pre-defined purpose, to collect and learn from a specific set of data for specific use cases (for example, a cybersecurity tool just for endpoint protection, or a digital marketing tool just for web analytics). This can be sufficient for specific knowledge or tactical decision making, but difficult to apply to strategic, complex decision making. This is because cross-contextual analysis is really manual and difficult, especially as the number of relationships between variables grow — again, leaving humans to connect the dots.
…And as we saw, human intelligence is not always reliable :)
Companies looking to aid human intelligence with machine intelligence could benefit from asking such questions:
- Am I getting intelligence or more knowledge from this service? Or is it forwarded information?
- How exactly is this service applying intelligence to its algorithms or models?
- How will the output be applied to improve my own knowledge or inform my decisions?
- Will we need to include additional processes to interpret the solution’s output?
How Human Managed applies intelligence
Extracting useful data from tools, examining alternatives through machine and human analysis, weighing up variables and discovering causal relationships is certainly possible. But the challenge is doing this consistently, at an organisational level, amidst a rapidly changing and ambiguous world.
Here is how Human Managed addresses this challenge: 1. Deconstruct 2. Collect & Process 3. Combine & Choreograph 4. Decision & Action 5. Improve
- First, we deconstruct the problem in the market from multiple layers: users’ needs, market players, available technologies and in particular, what the users are not getting despite the available players and technologies.
- We understand the difference between data, information and knowledge and only collect and process the relevant data, information and knowledge. We know what makes an output just ‘good to know’ versus ‘must know’. For example, we apply machine intelligence at the log level to ensure that our machine learns and processes at the same time. We apply human intelligence to understand our users’ specific problems and apply organic analysis on top of machine analysis.
- We combine and choreograph the tools, processes and interactions on our modular platform to produce relevant intelligence for particular users and moments. For example, the knowledge and intelligence will look different for analysts, engineers, executives and boards, or for operational situation versus a crisis situation.
- We co-create knowledge and intelligence for different scenarios with machines and humans (internal employees, customers, partners). The outcome is direct influence or assistance in our users’ decisions and actions. For example, our Workspace Orchestration Theme (Digital Story) resulted from applying our own knowledge as well as our tech partners’ and customers’ and it helps our users to take decisions on their remote workforce.
- Finally, we continuously improve on our data, information, knowledge and intelligence to improve our users’ decision making.
The future of intelligence is collective
Today, data and information are heavily commoditised. However, despite easier and cheaper data input, we do not get nearly as much intelligence to inform decision making.
Of course there are exceptions, but the process of interpreting data to intelligence remains largely separate between machines and humans in organisations: machines analyse large volumes of data through their models whilst humans analyse what they can comprehend through their own mental models. Multiply this by the number of business units and departments, each with its own army of vendor-driven tools… you start to see why people say they are not making decisions based on their data.
We believe that this is one of the biggest problems that the market faces today and it’s only going to get more pressing.
COVID-19 has been the biggest driver of Digital Transformation and will continue to be. Consumers and employees expect better experience, flexibility, interaction without compromising on security and privacy. Companies are now forced to face data head-on and decision makers have to make tactical, strategic risk-based decisions faster in all aspects of business.
Human intelligence and machine intelligence must come together to support a shift this big.
The great news is that the tools and techniques are already here — all that is required is a new way of thinking.
Whether you are a magician or a business leader; it does not matter what you know if you cannot apply the right knowledge at the right time.
That is intelligence on-demand.
* * *
If you found this content interesting, check out the Human Managed’s blog Human Thoughts, where we cover a wide range of topics that we are passionate about.
Stay on top of Human Managed’s latest updates including our backstage happenings on Instagram, LinkedIn and Twitter.
Originally published at https://www.linkedin.com.