PRODUCTHEAD: The good, the bad and the ugly of machine learning
PRODUCTHEAD is a regular newsletter of product management goodness,
curated by Jock Busuttil.
paranoid product manager
Machine learning is the science of getting computers to act without being explicitly programmed
Netflix and the BBFC automate age certifications for content with a staff-trained algorithm
YouTube’s recommendation algorithm has the unintended consequence of creating echo chambers
An algorithm widely used in US courts for predicting future criminality exhibits racial bias
Filmmaker Shalini Kantayya explores the impact of biased facial recognition algorithms on human rights
In 2004, I attempted to build an automated system to help me identify the prospective customers most likely to want to buy my products.
It relied on being able to link an individual’s different activities together, such as viewing product pages on our website, downloading a trial version, or contacting us in some way. It would then assign scores for those activities.
Crucially, I also wanted it to use a Bayesian algorithm, much like the one I used to filter my email spam, to gradually learn to boost the scores of the combination of activities that usually would result in a product sale.
Needless to say, I failed. (Otherwise I suspect my career would have gone in a somewhat different direction.)
While my ambition was unbounded, my technical skills and the technologies I knew how to use had more practical limits.
Access to technologies with this power bring with it an inevitable ethical responsibility, as I’ve discussed before. You can even now do a Master’s degree in the responsible use of artificial intelligence (AI).
So this week, I’ve pulled together some articles for you on the good, the bad and the ugly of machine learning.
Speak to you soon,
what to think about this week
AI is a branch of computer science attempting to build machines capable of intelligent behaviour, while machine learning is the science of getting computers to act without being explicitly programmed.
You need AI researchers to build the smart machines, but you need machine learning experts to make them truly intelligent.
Proteins fold up into complex three-dimensional arrangements, which in turn determine how proteins change its function.
Until now, predicting the folding arrangement has been a computational challenge that has resisted the efforts of biologists for decades, even with supercomputers at their disposal.
After training, DeepMind’s AlphaFold could deliver results in a matter of days, rather than the several years needed to examine proteins by hand, and averaged an accuracy of 92/100.
Netflix has labelled all of its content with a UK age rating generated by an algorithm.
The technology was developed with the British Board of Film Classification (BBFC), which classifies films in cinemas.
Netflix staff watched its entire catalogue, tagging sex scenes, depictions of violence and swear words, then fed this data into the algorithm.
Whether you’re new to product management or have been a product manager for years, a coaching session can help you to step up your career.
We’ve coached people wanting to get into product management, product people with nobody in their organisation to manage them, and experienced product managers preparing to apply for a promotion.
A proportion of the fees from every coaching session is donated to charity. Contact us if you’re interested in finding out more.
Guillaume Chaslot used to work on YouTube’s recommendation algorithm. He describes how their goal to increase the amount of time people spent on YouTube had unintended — but not unpredictable — consequences.
Modern society sits at the intersection of two crucial questions: What does it mean when artificial intelligence increasingly governs our liberties? And what are the consequences for the people AI is biased against?
CODED BIAS explores the fallout of MIT Media Lab researcher Joy Buolamwini’s discovery that facial recognition does not see dark-skinned faces accurately, and her journey to push for the first-ever legislation in the U.S. to govern against bias in the algorithms that impact us all.
Perhaps we’ve been caught a little off-guard by the implications of these new technologies. These have presented product managers with yet another new challenge to add to the growing list: how to create products that are not only successful but also ethical.
[I MANAGE PRODUCTS]
As individuals, we’re continually evaluating options and taking decisions. As product managers, we have the additional responsibility to balance the often competing needs of users, the business and wider ethical considerations. What makes one decision better than another?
[I MANAGE PRODUCTS]
A product is often a complex combination of several products and services. Some you create yourself, some are created by others. You’re responsible for the whole lot, even if they’re not all directly in your control.
[I MANAGE PRODUCTS]
So this this might seem a bit odd, but are there any lessons from antiquity given your background in Classics that one might apply to product management in the year 2020?
[I MANAGE PRODUCTS]
upcoming talks and events
5th May 2021, 16:00 GMT
Online product management round table discussion (topic TBC)
can we help you?
Product People is a product management services company. We can help you through consultancy, training and coaching. Just contact us if you need our help!
Helping people build better products, more successfully, since 2012.
PRODUCTHEAD is a newsletter for product people of all varieties, and is lovingly crafted from the ashes of failed food delivery startups.
Subscribe to PRODUCTHEAD
PRODUCTHEAD is my free curated newsletter of the best articles, videos and podcasts from product leaders and commentators all over the world. All neatly packaged up in a weekly email delivery for your reading, viewing and listening pleasure.
Read more from Jock
The Practitioner's Guide To Product Management
by Jock Busuttil
“This is a great book for Product Managers or those considering a career in Product Management.”— Lyndsay Denton