Getting there first: Startups or AI? Thoughts for founders at #SuS17 Edinburgh
I’ve been invited to Startup Summit 2017 in Edinburgh, to talk about how contextual changes, including rapid technological advancements, will change the world in 10 years time.
What piece of advice should I give budding entrepreneurs in front of me, in this gorgeous light-filled ballroom overlooking Edinburgh’s main street, about what they should prepare for on the 1st of November 2027?
Should I reassure them that tech giants will not develop AI so powerful that it will grab all opportunities from unsuspecting predictive and data analytics startups? Or should I encourage them to continue working for years on machine learning algos achieving super-human performance in lab tests, even if their access to live data-sets is restricted?
Prediction or precision in #AI — which is a worthwhile entrepreneurial pursuit?
With AI leading the way, one can question what constitutes a truly important development, and how should founders take advantage of it today, so that years from now they’d be thriving or at least… “safe”?
It’s not General AI that startups should worry about.
By 2027, founders will witness a breaking point in ICT and their markets, where computational power will transform knowledge work just like the steam engine broke the boundaries of physical labour.
What does that mean in terms of challenges waiting to be solved by today’s founders?
On one hand, narrow AI will easily outperform humans on high-level reasoning tasks, as raw computational power is expected to double every 18 months.
That’s a low hanging fruit , a quick win— to develop machine learning algorithms that augments, to a certain extent, cognitive tasks involved in pattern recognition and prediction, simply because computational power is getting cheaper and more accessible.
Hence, if you’re a budding entrepreneur in Edinburgh today, and want to leverage the power of AI in the best 5 years, the prediction and pattern recognition are easy wins. Grab data sets from Edinburgh Centre for Carbon Innovation or from the Edinburgh BioQuarter (via @scotent ), and apply machine learning, predictive modelling and dimensional reduction to drug discovery and landscape mapping.
On the other, hand, intuitive robotics and automation of low-level motor tasks will be more resource consuming and more difficult to crack. So that’s difficult, but is it worth it?
That’s Moravec’s paradox, in a nutshell – low level skills will require more computational power, whilst high level reasoning skills (prediction) will be commoditised as it requires much less computational power.
So if the building blocks of your startup look like information packages from complex data sets, then prediction is a quick win for the next 5 years.
If you are looking further into the future – and aim to stay relevant and successful over the next 10 years, I believe that personalisation and precision services, enabled by AI, will be the key to your success.
Think about it – if computational power becomes cheaper and more accessible (Moore’s Law), then most AI first startups will get into prediction, automation and pattern recognition.
If you want to differentiate yourself, recognise that there’s huge untapped value within providing deeper, more personalised services to customers. That’s possible in precision medicine or healthcare or in personalised financial advice.
I was very pleased to notice that one of the startup pitches at Startup Summit 2017 in Edinburgh was for personalised nutrition.
Here’s one of the founders, explaining how the value chain of building a machine learning startup which will eventually provide fine-tuned, deeply personal food consumption advice to individuals.
Her name is Tarryn Gorre (@tarryngore ) and she’s building @Kafoodle.
Tap into your values
Just days before my talk at Startup Summit, Future Advocacy had released a report highlighting which constituencies are at risk of automation. South Edinburgh is the least likely to be suffering from job displacement due to automation. Well done!
So what’s so special about Edinburgh that make it more resilient than any other city / region in the U.K.?
As it turns out, 40.1% of workers in this area work in jobs that require complex, high-level skills, such as healthcare and education. While wandering around chatting with attendees and exhibiting companies, I also noticed this beautiful, distinctive propensity towards meaningful entrepreneurship — the kind that isn’t just driven by profit, but by level of impact as well.
If that’s something specific to entrepreneurs in the North of Great Britain — I don’t know, but it’s certainly something to tap into.
As my co-panelist @BruceWalker_ from We Are The Future mentioned — it would be desirable to urge entrepreneurs to consider compassion, inclusion and social impact as business goals, on top of staying profitable, growing and expanding globally.
We Are The Future was also the driving force behind Startup Summit, and their goal is to help some of the bravest and most brilliant entrepreneurs out there make a difference.
In other words, as Adam Purvis said in his closing remarks — nurturing the most successful, self-sustaining entrepreneurial ecosystem in Scotland start with giving yourself permission to try, permission to fail, and permission to succeed.
It must be done thoughtfully, supporting old system in making the transition to whatever the future looks like; nonetheless, go out there and be the best (version of what you can be.
Special thanks to Adam Purvis of FutureX (@futurexglobal), Bruce Walker (
) of We Are The Future, and to Victor Alexiev (@victoralexiev) — Singapore AI Ambassador.The panel I was part of explored the topic of Startups in 2027. Facilitator: Adam Purvis (Future X), speakers: Josh Littlejohn of @socialbite_, Evelyn McDonald of Scottish Edge, Callum Murray of Amiqus, and Bruce Walker of We Are The Future.
