About us

Thrive is set up by Funding London, a venture capital company bridging the finance gap for early stage businesses based in London. With over a decade’s experience in supporting the startups of London through a variety of funding vehicles, Funding London sensed a need to illuminate the ever-evolving scenario of London’s early stage businesses.

Thrive features interviews with and opinion from budding entrepreneurs, investors and industry experts. A mix of contributors from all areas of the industry is desired in order to spark genuine discussion about ongoing critical issues. While it showcases the effectiveness of successful ventures, it also encourages sharing lessons learned from missteps and unsuccessful projects.

Visit Site

Credits

Content
Funding London
Design
dtc

Contact

info@thrive.london
020 7043 0739

First Floor, Aldwych House,
71-91 Aldwych House
London, WC2B 4HN

Follow

Perceptions · 1 May '18

When data science becomes a force for evil

It was inevitable. Our industry has had its first, really bad case of ethical rule breaking. And not just ethical, quite possibly legal as well. The story of the shady methods used by Cambridge Analytica in the run up to both the latest US election, and the Brexit vote in the UK, is still unfolding. It is clear that some unsavory practices took place, whether actually illegal or not remains to be seen. In some ways, legality is of minor importance in our industry, as regulators are at least two decades behind in legislating best practice in data science and AI. Our industry cannot rely on regulators and governments to take care of this problem, we need to do it ourselves.

But that is where the problem starts. As a regular speaker and panelist at data events, I have gotten the question on “how do we make sure we do nothing wrong?” countless times. My standard answer, which may need revision now, was always that “we will make mistakes, but the industry will self-regulate”. Customers will go elsewhere, employees will quit, brands will be destroyed. The latest news about Cambridge Analytica are not the first time they have been criticized, so why has the industry not rejected them yet?

I think the answer is two-fold. As long as there is shady money, there will be shady businesses. Hence, there can be customers who are even looking for those willing to bend the rules. That seems to be the case here. The second side of the coin, which is also where we, as a community, can influence the situation the most, is the individuals willing to do the work.

I happen to personally know several individuals who are currently, or have in the past been employees of Cambridge Analytica. They are normal people, the sort of people I would share a drink with on a sunny day. They are not inherently evil, or the sort of people who would commit offences.

My hypothesis is that when such curious, ambitious individuals who love solving problems and finding solutions join a company, they are given a problem (e.g. “how can we harvest Facebook profiles?”) and they will attack the problem straight on, without thinking much about whether it is right or wrong. Add in a toxic culture, where the best solution is celebrated and no one questions the purpose, and you have a dangerous mix of a highly skilled, incredibly intelligent workforce whose hard work can be twisted for bad purposes.

I also think diversity, or rather the lack of diversity, comes into play here. The data science industry and workforce is still very homogeneous. Without different backgrounds and viewpoints, it is easy for a data science team to end up in an “echo chamber” where certain opinions and practices are reinforced, and critical voices subdued.

So if we can’t trust the government or regulators to prevent these situations in the future, what can we do? Well, the data science community is small and well networked. We are a force to be reckoned with, if we team up. We can regulate ourselves! Ideas around some form of “Hippocratic oath” for data scientists, or a manifesto to which we should all adhere, have floated around in the community for a while. Let’s do it. Let’s agree on something, and sign up to it. Even if it is as simple as “do not to others, what you would not want done to you.”

I would also strongly encourage data scientists everywhere to become whistleblowers, if you see something you think is wrong or ethically dubious. We should “name and shame” companies that misbehave, and the word should spread like a wildfire. This would discourage echo chambers from forming, and make it hard for those businesses to hire a good team.

And we should work harder than ever to improve diversity in the industry. We need more women, people of different backgrounds and walks in life to join this profession and make their voices heard. What we need is more open and vocal debate.

Finally, we cannot do all the work ourselves. There will always be fringes to any industry that will do whatever they want. Hence, I do call for regulators to work harder and faster at creating up to date rules. I welcome scrutiny, as most of us will have nothing to hide, and any weeding out of bad businesses will only help the good ones even more. Governments everywhere, work with us to ensure an industry where good, ethical businesses will thrive and grow, and support a healthy ecosystem and society. Together, we can do it.

Note: This is a topic very warm to our heart at Pivigo, and we have written about it on our blog as well.

Disclaimer: I am not referring to any particular individuals currently or previously employed by Cambridge Analytica in this article, and none have been consulted about this text. The opinions voiced here are mine alone.