Do algorithms outperform us at decision making?
Hello everyone,
We face multiple decisions a day: What should I have for breakfast? What’s the best route to get to work? Which project should I prioritise today? Some decisions we face have high stakes, such as choosing a treatment for a patient or making a financial investment.
As humans, our ability to make decisions can be limited by biases and irrational thinking, so could we turn to algorithms to improve our decision making? We asked 6 experts in computer science, mathematics, and artificial intelligence (AI), ‘Do algorithms outperform us at decision making?’, here is what they said…
EXPERT CONSENSUS
Do algorithms outperform us at decision making?
5 out of 6 experts say ‘yes’
What is an algorithm?
An algorithm is simply a set of rules or calculations which takes input information and gives you an output. There are three major components to an algorithm:
The input data (e.g. the words ‘cake’, ‘apple’ and ‘banana’)
A process or calculation performed on this input data (e.g. sorting by alphabetical order)
The output data (e.g. the words in the order ‘apple’, ‘banana’, ‘cake’)
What are algorithms used for?
Algorithms are central to computer programming and data processing. They allow for information on the internet to be organised and accessible, and prioritise which hits come up first in search engines and on social media. Algorithms are not just confined to computer science but used in many fields, such as in medical settings to assist diagnoses or in the financial world to make economic forecasts.
Algorithms are increasingly being used to assist or, in some cases such as self-driving cars, completely remove the need for human involvement. Their usage permeates every aspect of our lives, from matching us with a potential partner on a dating app to planning our route to work.
Do algorithms outperform us at decision making?
Professor Scott Fahlman, an expert in computer science and AI from Carnegie Mellon University in the USA, says “In many domains, yes. For example, chess is a decision-making process, and AI-based systems can now beat the world champion.” The AI algorithm Deep Blue beat the world chess champion Garry Kasparov in 1997.
Dr David Tuffley, an expert in software science from Griffith University in Australia, highlights a caveat, saying “A well written algorithm does have the potential to outperform the notoriously quirky and inconsistent decision making that many humans perform. But it really depends on how well it was written by a human in the first place. They just do what they are told to do.”
Professor Mark Lee, an expert in computer science from Aberystwyth University in Wales, says “Consider programs that play games like chess or Go. These AI systems have now reached grandmaster level and have beaten the world's best human players. The search algorithms used in chess can analyse millions of possible moves in seconds and this allows them to see much further ahead in the game than humans can manage. This produces excellent decisions and many game playing programs can now learn to play a game from scratch and quickly become champion players.”
Are there situations where humans make better decisions than algorithms?
Professor Lee says “games are a very constrained kind of problem. In a board game everything is well defined and consistent – nothing else happens other than pieces are moved! In the real world, in our human environment, there are all kinds of problems for algorithms to overcome.”
He goes on to say “Just deciding where to take a holiday, for example, can be influenced by weather forecasts, geographic choices, travel convenience, personal preferences, interactions with family and others, calculations of acceptable budgets of time and cost, etc. These are not straightforward yes or no issues, unlike making a move in chess, but involve uncertainty, missing information, and value judgements. Algorithms that attempt to replace general human decision making have not been successful. AI research is working in many areas that help us to make decisions, perhaps the most useful is in providing relevant information, but the subtly of the human mind still beats machines in many ways. So, yes, algorithms can massively outperform humans, but only in certain limited contexts and restricted applications.”
Professor Fahlman agrees, saying “Current AI systems cannot yet match human abilities where broad, common-sense knowledge is required, or when dealing with unpredictable, complicated humans.”
Professor Kay Kirkpatrick, an expert in mathematics and AI from Illinois University in the USA, highlights something humans have that algorithms don’t: emotions. She says “It’s recently becoming well understood in cognitive science that emotions can help humans make decisions well and quickly, and there’s nothing comparable to this in algorithms so far.”
The takeaway:
Algorithms can outperform human decision making, but currently only in certain contexts.
May the facts be with you!
Eva
Can AIs forget?
Learn more about the AI with our deep-diving review.
If you like our newsletters you can make a donation to support our work here at Metafact.
Verified by Metafact is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.