Two-faced tech

# The early days

I’ve been optimistic about the benefits of technology ever since I got my first computer in 1981. I’ve owned a ZX81, VIC20, Commodore 64 and BBC. A Sinclair ZX81 cost £70 in 1981, equivalent to £275 in today’s money¹. My father was generous, spending on his 11-year-old son’s hobby, and I am grateful he did, because it’s the reason I became an engineer. Wind forward to 2014, and my faith in technology, or how we humans use technology, has been dented.

It wasn’t long before my early computers led me to dial-up modems and a pre-internet service called Prestel. From the comfort of my bedroom, I could communicate with hundreds of like-minded people using bulletin boards. It felt revolutionary, even to this 11-year-old. Prestel was launched by the Post Office in 1979. It never quite achieved the success it could have, but it was an important stepping stone on the journey to today’s internet and online collaboration.

Multiple inventions made the internet possible, not least TCP/IP protocols, the brain child of Vinton Cerf, Bob Kahn, Bob Braden and Jon Postel². ARPANET, a military project, created TCP/IP to link computers across the world. Contrary to popular belief, ARPANET’s motivation in the 60’s wasn’t to create a network to withstand a nuclear attack, but rather allow remote “time-sharing” of scarce computer resources.

I sent my first email in 1991 at uni. Granted, it was only to someone in the next building, but it was mind-blowing knowing a message arrived at its intended destination within seconds of being sent. We take this for granted now, but we’re talking about an era of physical letters.

More recently, Artificial Intelligence (AI) and it’s sub-branch, Machine Learning (ML), have caught my attention. AI is:

“…the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence….”

• John McCarthy 2004³.

AI solves problems that were once the preserve of humans. ML uses algorithms and data to mimic human learning, enabling it to complete tasks without explicit instructions. Benefits AI and ML have delivered include:

• Google’s Deep Learning machine learning program is accurate 89% of the time in detecting breast cancer. That’s much more effective compare to pathologists at 74%. (Google AI Blog)
• The accuracy of Google AI’s machine learning algorithm in predicting a patients’ death is 95%. (Bloomberg)
• Using a machine learning algorithm, Amazon automates picking and packing items in a warehouse logistics setting. With Kiva’s capabilities, Amazon’s average ‘click to ship’ time reduced by 225% from 60-75 minutes to 15 minutes. (Forbes)
• Netflix saved $1 billion in 2017 as a result of its machine learning algorithm, which recommends personalized TV shows and movies to subscribers. (Forbes)

These technologies have been game changers, but there’s a flip side, something we engineers often overlook.

# Spam

The first email was sent in 1971 and by 1996, email overtook postal mail⁴. But, by 2003, spam accounted for more than half of all emails, rising to 90% by 2010.

# Addiction and lies

The first “social media” platforms appeared in 1976, called bulletin boards, a term coined by the tech world to describe applications that let you could post and read comments. Bulletin boards were only available to people with access to a computer, and the technical know-how, which was a tiny population. On the whole, these communities were mostly productive environments for collaboration.

Wind forward to today, the majority of the planet has access to 24/7 social media platforms, but how these are used isn’t all good. No one can deny the harms of social media. FOMO (feeling of missing out) aid compulsive behaviours, where people are constantly checking their accounts and zoning out from the real world around them. This is deleterious to mental health, and the anti-thesis of being social.

Social platforms can be echo chambers of bias, where conspiracy theories and fake news reach millions in a heart beat. The multiplier effect of these networks is harming people and threatening democracies at an unprecedented speed and scale⁵.

And if this wasn’t bad enough, you have this depressing statistic….

! [[ teenage girl suicide rate due to social media.png ]]

# Automated amplification

ML algorithms “learn” what interests, specifically what gets your attention. Armed with this information, they fill your feed with more of the same, so you come back more often, and stay for longer — it’s called engagement. It’s the reason you can scroll your day away on your phone. It doesn’t matter if this engagement is positive or not, all that matters to the social media company is getting engagement by giving you your “hit”. If you believe in COVID-19 conspiracy theories, you’ll get more of the same, reinforcing your beliefs. Like a magnet, it brings together like-minded, but misinformed individuals. Instead of expanding your knowledge and exposing you to different opinions, the opposite is happen. It’s polarizing, and that’s dangerous.

# Replaced by a computer

By 2025 AI could replace 85 million jobs around the world⁶. The creation of new jobs will likely offset most, if not all, of the losses, but the new jobs will require training, or retraining, people. This requires long term investment, particularly in education. Judging by our govt’s track record in this kind of investment, I won’t be holding my breath waiting for the investment to arrive anytime soon. If you think it only affects low skilled jobs, think again⁷⁸.

The societal impact of AI technologies needs urgent debate. The UK decimated entire industries in many northern cities, such as coal mining and ship building, leaving large swathes of communities impoverished and disengaged from politics. This caused understandable friction, and whatever your views on Brexit, there’s no doubting it influenced voting. It’s left the country divided in a way that I’ve not seen before.

# Summary

I don’t know if my concerns are overblown. What I do know is I’m less optimistic, at least in the short term, about the impact of technology, than I once was.

# References