Capital Enterprise November AI Round-Up

Published in

Capital Enterprise

10 min readNov 14, 2017

Hello! This is our second blog post in a series. This publication will explore and provide input in the AI tech scene, highlight interesting AI startups, and speak about tech events of note.

Building an Artificially (More) Intelligent UK through startups, by John Spindler

The UK is one of world’s Artificial Intelligence (AI)/ Machine Learning epicentres. This has arisen not only because we have 4 of the top 20 Computer Science Universities in the world (Cambridge, Oxford, Imperial & UCL) generating an amazing pipeline of talent, but also because the UK AI eco-system keeps on generating amazing startups. Other leading AI nations have built their position at the forefront of the so called AI powered “4th Industrial Revolution” on the back of their dominant market positions and investment of their leading tech companies, such as Google, Amazon, Microsoft, Facebook and Apple in the USA, and Alibaba, Tencent and Baidu in China. In addition, these nations, as well as those such as Canada, Singapore and France, have benefited from significant investments from their governments to support research and the development of their AI powered economies.

In comparison, the UK government’s commitment to AI so far has been much more modest, and few people will be surprised that none of our leading Footsie 100 corporations are known for their investment in AI or Machine Learning. That the UK is renowned as a centre for AI and Machine Learning is not only due to our amazing universities, their amazing academics and students, but most importantly, our amazing startups. In short, this is the history of what has happened:

Step 1. Academics, researchers and students from UK’s leading universities have in the past 10 years created AI startups that have gone on to build deep world leading AI technology, meaning they were bought relatively quickly by predominately big, American tech companies.

This includes startups such as Vocal IQ (backbone of Apple’s Siri), Evi (the creator of Amazon’s Alexa), Swiftkey (acquired by Microsoft), Magic Pony (acquired by Twitter) and the big daddy of them all, DeepMind (acquired by Google, and now after the Alpha Go success, the world leading private sector AI research centre).

Step 2. The high profile success of these AI startups/research projects has enabled our universities to attract some of the world’s best academics and students to the UK, which in turn has led to the world’s leading tech companies (Microsoft, Amazon, Apple, Samsung etc) to locate AI research facilities to the UK (often built around the AI companies they purchased)

This means career AI researchers can now not only carry on researching and publishing papers as a university employee, but also now in the UK can carry on being a researcher as a much better paid employee of a company like Google’s DeepMind.

Step 3. The talent and examples of entrepreneurial success has spurned a minority of the AI students and academics to either start or join the next wave of amazing AI startups.

Companies like Improbable (in Games development), Benevolent AI (in Drug Discovery), Five AI (in Autonomous Vehicles), Graphcore (in Microchip design) and Onfido and Ravelin (in Finance). Some of these startups benefited from going through one of the UK’s world leading tech accelerators that specialise in the sector, such as Entrepreneur First, Techstars and more recently, Deep Science Ventures.

I am proud to say many of these universities and accelerators are Capital Enterprise members, and I can confirm if anything the pipeline of potentially great AI startups is getting stronger.

Step 4. In turn, this pipeline of new startups (and the knowledge that there are international companies looking to acquire UK AI startups) has attracted investment into the sector.

A recent report by MMC Ventures on AI investments found that in the USA and Europe, 18% of all the AI startups that had raised VC money were located in the UK. In no other sector, not even Fintech, does the UK have such market share.

Step 5. The above factors are creating a network effect, especially in and around London, that is attracting and enabling talent to form startups that can attract business and technical expertise and support, enabling these startups to build out commercial AI applications that can secure POC pilots with early adopting clients.

This in turn enables them to establish the “proof points” that attract investors and investment, leading to more startups and the fly-wheel repeats.

AI in the UK, especially London, is startup led. A fact down played in the otherwise excellent Government commissioned independent 2017 review, carried out by Professor Dame Wendy Hall and Jérôme Pesenti, reports on how the Artificial Intelligence industry can be grown in the UK.

In our view, helping startups acquire the resources, talent and investment needed to build AI/Machine Learning powered game changing startups is how the UK can cement its position in the “4th Industrial Revolution”. The government can help.

Ben Medlock, the co-founder of Swiftkey, revealed at last month’s World AI Summit in Amsterdam that it was a small government innovation grant that launched Swiftkey. So everything helps. This is why Capital Enterprise is doing its small part by launching two new initiatives:

CAP-AI

A partnership between Capital Enterprise, The Digital Catapult, Barts Health Trust and the Alan Turing Institute, this two year project will seek to foster and expand collaborations between startups and academia.

The Digital Catapult will be building a Machine Learning Lab that enables startups to access a private GPU powered cloud, and therefore help startups not use all their AWS/Azure/Google credits on building and testing a prototype.

The Digital Catapult and the Alan Turing Institute will be bringing together UK leading academics, startups and industry players to develop applied AI solutions to the big issues facing UK corporates and public sector institutions.

Capital Enterprise will be co-funding PhD and post doctorates from UK’s leading universities to work on agreed R&D projects within London’s leading AI startups, and much more besides to help over 125 London AI startups obtain the latest research, best talent, and resources to build out their cutting-edge AI business.

The programme is co-funded via the European Regional Development Fund, but Capital Enterprise is hopeful that once CAP-AI establishes a successful track record that it could obtain government and private sector funding, and thereby could be rolled out across the UK. CAP-AI is being launched on the 21st November at the Digital Catapult Centre, at an event sponsored by Google Cloud. If you would like to attend please contact gracie@capitalenterprise.org.

AI Seed

Capital Enterprise has partnered up with a new fund, AI Seed, to find and support UK based AI startups. Each financial year AI Seed will invest £100K for around 5% of the equity into very early AI startups when they emerge from universities and accelerators.

AI Seed is willing to lead seed size rounds, find and bring in co-investing specialist angels and VC’s, and provide extensive post-investment support. Most importantly AI Seed has the expertise to carry out in-depth Technical Due Diligence and therefore can invest on the potential of the team and technology and not just on market traction.

So far in 2017–18 AI Seed has invested in 10 startups, including:

Transformative AI — A team from Cambridge University that is seeking to save lives of at risk patients by giving them up to 20 minutes warning of a pending Cardiac Arrest.
Observe Technologies — A team from Imperial College, who apply computer vision and machine learning to ensure fish farmers won’t miss any opportunity to maximise stock value.
Predina — A team that graduated from Entrepreneur First and are predicting the probability of accidents for dangerous goods haulage firms and helping them map out safer routes.
AI Build — A team that has graduated from Seedcamp and UCL, AI Build is developing 3Dp technology for the 3D printing of large scale objects cost efficiently at high speed.

All the startups Capital Enterprise are supporting are trying to use AI / Machine Learning to solve real, impactful problems that will make the world a better place. This is important to note for much of the discussion in the media about AI paints a picture of a coming dystopia where AI powered robots will replace humans and most of us will be unemployed. There is very little evince (as opposed to visionary speculation) that this will ever happen, whilst we are missing the real picture:

If we do more to help AI startups in the UK prosper, they will produce tools and solutions that will make us safer, live longer and act more intelligently.

A wonderful world powered by AI, built and deployed by UK based AI startups.

2. This month’s best AI blogs, by John Spindler

“What the UK can do to build world class AI Industry”, Independent report
“How an AI Start-up can take on the big tech giants and win”, by Louis Coppey, VC at Point Nine Capital
“Why AI is at an “Inflection” point”, by MMC Ventures
Quarterly AI Review, by Nathan Benaich — A technical but brilliant must read for anyone interested in what is happening in AI research and startups
Weekly review of everything in Deep Tech, by Azeem Ashar, Exponential View — Amazing Sunday morning read
Data Science, AI and Machine Learning Briefing, by CognitionX

3. Our top 5 AI startups to watch

Evolution AI

Their technology reads, understands and processes text documents autonomously, such as web pages or scanned documents, freeing up staff for higher-value work. Evolution AI are certainly a hot startup to watch.

Flux

Flux, the receipt data platform, have recently closed a $1.5 million seed round led by PROfounders, the investors in made.com, among other high profile investors. The funding will be used to hire across key roles, particularly engineering, in order to support Flux’s fast growing platform.

Personalyze

Personalyze is an award-winning AI-driven people analytics company, that has created the world’s most powerful profiling technology on individuals across social and mobile, enabling us to understand people better than anyone else. They have recently raised £1.27m, in which Capital Enterprise was involved.

FactMata

FactMata have recently raised $750,000 (£575,000) in seed funding, with an aim to use artificial intelligence for automated content scoring and verification, and to tackle fake news. The firm raised funding from American businessman, investor and television personality Mark Cuban; Mark Pincus, the founder of Zynga; Ross Mason, the founder of enterprise software company MuleSoft; and Sunil Paul, founder of anti-email spam software company Brightmail.

Prowler.io

Prowler.io recently raised £10 million ($13 million) to use AI /Machine Learning to focus on behavioural learning and simulation in virtual environments. Led by new investor Cambridge Innovation Capital, the round also had participation from Atlantic Bridge Capital as well as previous investors Passion Capital, Amadeus Capital Partners and SG Innovate.

4. AI Events of note in the next month

SLUSH

On the 30th November -1st December, Slush will bring together high profile actors of the global tech scene in Helsinki for the world’s leading startup event.

The event aims to help the next generation of great, successful companies come forward, and has grown in just a few years to 17,500 attendees and 1 million live stream viewers.

NIPS 2017 — Conference on Neural Information Processing Systems

The 31st Annual Conference on Neural Information Processing Systems (NIPS), taking place on 4th-9th December 2017 in Long Beach, is a multi-track machine learning and computational neuroscience conference that includes invited talks, demonstrations, symposia and oral and poster presentations of refereed papers.

CAP-AI Launch

We are proud to be launching our newest programme, CAP-AI, on Tuesday 21st November at the Digital Catapult Centre, from 6–9:30pm. As mentioned above, CAP-AI is designed to help connect startups/SMEs with AI talent, to work on 6–24 month AI/ML related research and innovation projects on a funded basis.

The launch features a pre-event from 4pm where Google will be holding a series of panel discussion and demos. The official launch of CAP-AI starts from 6pm, featuring panels, speakers, and the opportunity to network afterwards.

By John Spindler, Nick Slater and Gracie Jones