Cambridge Consultants tackling innovation before the curve has arrived
Cambridge Consultants held its first Innovation Day at its recently-opened Science Park extension – and gave visitors from around the globe some jaw-dropping hints about future technology soon due to change the world around us.
“It’s three years since our last Innovation Day,” said Cambridge Consultants’ CEO Eric Wilkinson, addressing the invited audience of current and previous clients, “though we’ve had innovation days in Boston, San Francisco and Tokyo – Japan was our third biggest market in 2018. In that time we’ve invested further in this site, this building opened in May, we have wet and dry labs here, it’s a design hub, with food facilities for 1,000, a rooftop bar and 200 in-house innovators.”
Mr Wilkinson showed a vial of liquid on the giant screens facing the audience. The vial contained the entire contents of Wikipedia, encoded as DNA by Catalog, one of Cambridge Consultants’ clients.
“Why did Catalog imagine they could write DNA like this?” asked Mr Wilkinson, though everyone knew it was because Cambridge Consultants had the wherewithal to develop any number of audacious technical feats. “We looked to the future, one of the options was synthetic biology, so we invested ahead of the curve.”
Investing ahead of the curve - the next step for the DNA technology is to improve read/write times - was very much the theme of the day and the first keynote, Tim Ensor, director of artificial intelligence for the company founded in 1960, said: “We’re entering a period where machines are starting to make complex and potentially critical decisions on our behalf, so it’s important we understand what AI is.
“The definition in 1955 by McCarthy, Minsky, Rochester and Shannon said that it is ‘some way of making machines behave in ways that would be called intelligent if a human was doing it’ and today, there’s a new focus on mathematics and algorithms, with the new wave of AI driven by machine learning.”
Tim showed one of the outputs of AI: Mona Lisa talking. The animation was startling, but AI is more than just an attention-grabber. Behind that shop window Cambridge Consultants has developed a digital pathology platform for a client using AI to monitor TB, “so doctors don’t have to do the work”.
Research is ongoing but early results suggest the platform is more accurate too.
“To extract the next wave of value we have to unblock challenges,” says Tim. One challenge that AI is solving is unveiled in the demonstration area. Luke Smith, a senior engineer in AI on the site, is part of a team of a dozen which has developed SharpWave, which resolves poor quality or damaged images using machine learning.
“We’re developing bespoke deep learning solutions for clients with algorithmic problems,” he says.
The SharpWave demo is an AI program presented with a very fragmented image which is then reassembled into a fully accurate image in real-time.
“The day is going well,” says Luke. “There’s been lots of conversations, especially about the underlying technology – we’re showing the relevance of this technology not just to images but audio and ultrasound too, to integrate into medical solutions and other applications.”
Along the corridor, Sally Epstein, strategic technology, Cambridge Consultants is doing a demo on what quantum computing can achieve.
“We invest in these new areas and communicate with really ambitious clients,” explains Dr Epstein, who was previously at Darktrace.
“We’re interested in software not hardware – what can you run on these quantum computer systems. I’d say it’ll take 30 years for purposeful quantum computing to be widespread and accessible to most companies.”
General-use quantum hardware may be decades away, but the software is already viable. A demo shows 100,000 films being scanned for a word or phrase in less than six seconds, using a quantum-inspired algorithm on one of today’s classical computers.
“Our team of visionary technologists has seized on the 2018 discovery of a quantum algorithm that will run happily on a classical computer,” Dr Epstein says.
“They’ve implemented their own version of this lightning-fast algorithm as part of their research into solving previously intractable problems for our clients. This quantum-inspired demonstration showcases one example – a recommendation system for a movie service provider.
“Calculations that we would expect to take hours, days or even weeks are delivered in real-time.”
One of the technical challenges Cambridge Consultants is looking at is to scale up quantum processing power, which is measured in qubits. Today’s computers manipulate bits that exist in one of two states: a 0 or a 1. Quantum computers aren’t limited to two states: they encode information as quantum bits, or qubits, which can exist in superposition, ie can be one and zero simultaneously.
This allows multiple calculations to be performed at the same time. Millions of them. Google’s 54-qubit Sycamore computer has performed a calculation in 200 seconds that would have taken the world’s most powerful supercomputer 10,000 years.
“What if Google’s Sycamore quantum computer could be scaled up to 5,000 qubits?” muses Sally. “There’s no guarantee it would work, but they’re all trying.”
The technical challenge is that “it’s very hard to keep stable, it’s very susceptible to magnetic fields”.
This very sensitivity makes for very good cybersecurity networks, as those attending the afternoon seminar on quantum superposition and entanglement learnt.
“Although we understand a lot about the universe at scale, at the sub-atomic level things can become counter-intuitive, impenetrable and, for some, mystical,” said Sally. “We’re ahead of the curve. We don’t know what the future will hold but we know it’ll be disruptive, but where will it lead? Drug discovery, new materials, enzyme discovery, protein discovery – it means the ability to design drugs according to a specific criteria.”
One of the ways quantum understanding can be utilised is cybersecurity: any intruders’ magnetic waves, which are inevitable in any attempt to penetrate closed systems, can be quickly detected.
“Quantum principles such as superposition and entanglement are set to solve intractable real-world problems,” says Cambridge Consultants of its interest in the sector. “Already, they’re being harnessed to guarantee eavesdropping-proof communications and scan inside the human body with unprecedented clarity.”
There will be further applications, says Monty Barlow, Cambridge Consultants’ first head of strategic technology.
“A quantum magnetometer can measure magnetic fields, for instance in an MRI scan.
“There’s also the option of better psychological understanding of people using half psychology and behavioural science techniques and half artificial intelligence. How that might work is, if you take Alexa, it doesn’t understand if your dementia is proceeding, so that’s a whole new human/machine understanding, that’s called HMU.”
I google it and find Monty’s HMU is indeed so new that Google hasn’t heard of it.
“There’s no name because it’s a whole new field!” says Monty. “Imagine building a team for a sector that doesn’t have a name yet.”
There’s other talks and demos on synthetic biology, healthcare innovation, and some really cool stuff involving VR headsets – one a cricket game, another involving the control of robotic hands developed for post-stroke patients.
The final panel also threw up some areas of interest.
On the panel were Jonathan Sackier, chief medical officer of neurotech company Helius Medical Technologies; Carl Nagle, global marketing innovation director of beverage company Jacobs Douwe Egberts; and Neill Hunt, executive director, innovation at synthetic diamond company Element Six.
During this ‘World of Opportunity’ discussion, Mr Nagle explained how coffee scarcity is pushing the development of “molecular coffee which is in every respect a facsimile of the coffee bean, but it’s been made in the lab”.
Dr Sackier referred to the need “to literally come up with a whole new healthcare transaction” and explained that some of this is patient-led.
“A lot of people like having their chemotherapy at home now,” he said. “Innovation drove the change.”
My main takeaway from this innovation showcase?
That when a cyberattack takes place, new types of instrumentation allow the organisation under attack to intercept the intruder at the photonic level in real-time.
Talk about ahead of the curve.