AI & Tech Update

AI investment in the UK is at record levels. Fintech is booming even as Facebook’s digital currency, Libra, stalls. Oh yes, and Google has just attained Quantum Supremacy, writes Victor Hill. What should investors make of it all?

The race for artificial intelligence intensifies

A report by Tech Nation released last month showed that companies working on international intelligence (AI) in the UK secured a record $1 billion of new investment in the first half of this year. This makes the UK the world’s third largest market for AI investment behind the USA and China.

Funding for H1 2019 exceeded total investment for the whole of 2018 and is six times that of 2014. In this respect UK AI research is well ahead of France, Germany and the rest of Europe – despite Brexit. The UK is surpassed only by the USA in the number of companies working on AI but note that 89 percent of UK AI outfits are small with fewer than 50 employees.

The concern amongst outsiders is that while these UK AI firms are incubating some incredible technology, very few – if any – will reach escape velocity to become a next-generation Alphaber/Google (NASDAQ:GOOG) or Facebook (NASDAQ:FB) with global clout. Moreover, there is a brain drain going on as Silicon Valley companies raid our top universities by offering packages to the best talent that cannot be refused.

Google’s purchase of DeepMind Technology for $400 million in 2014 is a cautionary tale. It pledged to preserve the London-based AI pioneer’s autonomy and to set up an independent board of ethics. But after Google was restructured under the umbrella of Alphabet everything changed. In August Alphabet/Google placed one of DeepMind’s founders, Mustafa Suleyman, on indefinite leave. This destroyed any remaining pretence that DeepMind is an autonomous entity.

Patrick Pichette, Google’s former CFO, was the Canadian who orchestrated the Google takeover of DeepMind. He now actively advises AI start-ups against selling out early. He wants to help UK AI incubators to build themesleves into global champions by securing more capital and coaching management to adopt global ambitions. This year he opened the London office of Canadian venture capital firm Inovia Capital. According to one report, Inovia has a war chest of $600 million to invest.

Climbing the Tower of Babel

AI and machine learning are accelerating the advance of translation technology, thus allowing people to communicate instantly with one another without any language barrier. Yesterday Vbestlife Smart Translation Bluetooth Wireless Earphone, which can translate instantly via an earpiece, was available on Amazon for £18.05 to UK Prime customers. Amazon and Microsoft both have their own translation engines which can be used to translate text almost instantaneously, while Apple has iTranslate. Google Translate’s camera function enables travellers to take snaps of signposts in foreign languages with their smartphones and learn what they say.

For almost the last century English has been the world’s dominant language – a position reinforced by the digital age which started in the early 1990s. According to a recent report by the University of New South Wales, English speakers still dominate the internet – but only just. Mother tongue English speakers account for 28 percent of global internet users; but Chinese speakers are catching up with 23 percent. Spanish speakers account for 8 percent. Yet, when it comes to total online content, 56 percent is in English, 5 percent in Spanish and just 3 percent in Chinese.


There is, apparently, huge demand in China for apps which can translate text from English to Chinese. Chinese is a tonal language (with four tones – Thai has five) while English is not (though it is stress-timed). Rendering English tones which denote emotion into the correct Chinese word carrying the correct meaning-tone has proved challenging. But neural networks which mimic human brains in order to teach themselves how to perform tasks are making progress on this.

Libra stalls

In the August edition of the Master Investor magazine I set out why I think Facebook’s proposed digital currency, Libra (announced on 18 June), was likely to be a game-changer. But over the summer eight out of the 28 companies and organisations which were to form the Libra Association have pulled out. The eight quitters include global payments platforms Visa (NYSE:V), MasterCard (NYSE:MA) and PayPal (PYPL).

What has gone wrong? Basically, the European regulators have taken a very dim view of the scheme and the US regulators have not exactly been enthusiastic either. The main issues are twofold. Firstly, governments fear that the cryptocurrency available through the world’s largest social media platform could be a godsend for criminals, money-launderers and terrorists. Secondly, the Europeans are concerned that it could undermine monetary policy and even the entire banking system. As a result MasterCard and partners have been spooked that they might encounter regulatory and fiscal sanctions if they proceed with the project.

In September, French authorities announced plans to block the roll-out of Libra in Europe, citing potential systemic financial risks and the possible abuse of Facebook’s market dominance. French Finance Minister Bruno Le Maire even claimed that the proposed cryptocurrency would undermine “government sovereignty”. He told an OECD meeting on 12 September that France “could not authorise the development of Libra on European soil”. Even in India, where Facebook subsidiary WhatsApp has over 200 million users and whose Prime Minister, Narendra Modi, is on cordial terms with Mr Zuckerberg, the reception was unenthusiastic.

On Wednesday (23 October) Mark Zuckerberg, Facebook founder and CEO, testified before the Financial Services Committee of the US House of Representatives, with the discussion focusing on Libra. Committee members questioned why Facebook would want to create its own digital currency, and why the company should be trusted in view of concerns about its exploitation of customer data.

Mr Zuckerberg’s defence was that Libra was essentially a payments system which would enable ordinary people to pay money to one another as easily as they currently send each other messages. He admitted that Facebook itself was part of the problem in so far as controversy about its use of customer data has undermined its credibility as the sponsor of a global cryptocurrency. The Libra Association, he reminded congressmen, would be an independent entity domiciled in neutral Switzerland.

But his killer argument struck home. America must embrace Libra, Mr Zuckerberg argued, in order to retain its financial and technological supremacy. If America blocks Libra then the Chinese could launch their own digital currency which might become the dominant global standard with negative consequences for American influence and values. The hegemony of the mighty US dollar might be lost overnight. We can assume that Mr Trump’s ears pricked up when he heard that. China’s central bank is known to be near to issuing its own cryptocurrency and the country is already ahead of the US in terms of payments platforms.

Under questioning by representatives, Mr Zuckerberg said Facebook would ultimately withdraw from the Libra Association if the cryptocurrency can’t earn approval from regulators. It is doubtful that if Facebook were to withdraw from the project that the other members of the Libra Association would wish to pursue the idea. Originally, the 28-member association said it hoped to reach 100 or more members when Libra is launched.

Increasing scepticism around Libra seems to have had a knock-on effect on Bitcoin, the well-established and more famous (if not notorious) cryptocurrency. On Wednesday it fell from £6,200 in value to £5,800.

I doubt that Mr Zuckerberg will give up on Libra. He will retreat symbolically – but in the frenzied underworld of Menlo Park, like the Nibelungen in Wagner’s Ring Cycle, the dwarves will continue to hammer away…

Google achieves Quantum Supremacy. Really?

On Wednesday (23 October) researchers at Google announced that their quantum computer called Sycamore had solved a problem that would have taken the very best conventional machine thousands of years to crack[i]. This milestone has modestly been hailed as quantum supremacy. Despite the hype, it represents just one more step, and a major one, towards realising the extraordinary potential of quantum computing.

Quantum computers are devices which exploit the strange properties of quantum physics to speed up calculations. As I explained in a previous piece on the subject, the smallest addressable components of conventional computers are bits which are either ON OR OFF. Quantum computers – in a nutshell – work with bits that can be on or off – or neither or both…Out of this weirdness, where each bit can exhibit four possible states, comes immeasurably faster computing capacity.


The scientific paper backing up the research, published in the journal Nature on 23 October, comes a month after a draft was accidentally posted on a NASA server. It demonstrates that a quantum processor consisting of 54 superconducting quantum bits – or qubits in the jargon – was able to perform a random sampling calculation (essentially verifying that a set of numbers is randomly distributed) exponentially faster than a conventional supercomputer.

Google’s quantum computer completed the calculation in just 3 minutes and 20 seconds, though one of the qubits had to be turned off as it wasn’t working properly. The version of the paper in Nature appears to contain no significant changes from the one leaked from NASA. It stands by the claim that the same calculation would have taken IBM’s Summit, the world’s most powerful supercomputer, about 10,000 years.

IBM (NYSE:IBM) then came up for air, claiming that, with some new algorithms, Summit could have solved the problem in just two-and-a-half days[ii]. IBM, which has its own 53-qubit quantum computer, prefers a higher threshold for quantum supremacy, and argues that Google has not yet reached the milestone. However, even if that is true, by my reckoning, IBM has tacitly admitted that Google’s machine is 1,080 times faster than its own.

Google has been working on quantum computing for 13 years, according to CEO Sunder Pichai. So has IBM, along with others like D-Wave Systems headquartered in Vancouver, Canada. D-Wave talks about Practical Quantum Computing on its website, along with quantum computing apps for things like determining airline schedules, chemistry simulations, and automotive design.

The paper will no doubt be scrutinised by computer scientists all around the world for months, if not years. No one thinks that quantum computing has arrived and will become ubiquitous overnight. However, as New Scientist writes this week, this is proof of a concept. In particular, the accuracy of Google’s machine is in doubt as there is suspicion that the system may have accumulated billions of tiny errors over the course of the entire calculation.

Non-specialists would be wise to assume that quantum computing will happen – though not overnight. It’s likely that weshall have to become more intelligent just to be able to apply quantum computing – just as you would not give a toddler a racing car to play with. In the first instance I can imagine that quantum computing could be used to take encryption to new levels of sophistication and complexity in the struggle for cyber-security. And perhaps, eventually, quantum computers may be able to solve the greatest challenge of theoretical physics: namely, how quantum theory itself can be reconciled with Einsteinian relativity.

Not that we would understand it.

Break up Big Tech!

2020 Democratic presidential contender Senator Elizabeth Warren has emerged as the nemesis of Big Tech. Senator Warren has accused the tech titans of abusing their market dominance and of crushing competition. She says that they have become monopolies which have “bulldozed competition, used our private information for profit and tilted the playing field against everyone else”.

The irony is that Silicon Valley is peopled with the youthful graduates of liberal-leaning East and West coast universities who champion progressive causes – apart from breaking up Big Tech, that is. Hitherto, they have been big donors to the Democratic Party; now, deep-down they think they might be better off under Trump II. And it’s not just Ms Warren. Bernie Sanders has lambasted Amazon over low wages and tax avoidance, while Amy Klobuchar has called for draconian new privacy laws.

The case against Big Tech has two main legs. First, they are accused of offering their own products on platforms they control. For example, Apple uses its App Store to promote its own software and Amazon uses its online retail portal to promote its own-brand products. Second, they take out competition by buying them up – just as Facebook digested WhatsApp.

The US presidential race has almost exactly one more year to run. We can be confident that the issue of regulating the tech titans will become a major theme in the race to the White House. But the irony is that Mr Trump, who personally loathes most of the tech super-billionaires and accuses Facebook of censoring conservative views, has consistently sided with Silicon Valley under the banner of America First. He even claimed that the EU had taken advantage of the US after it imposed a €5 billion fine on Google for anti-competitive practices. Mr Trump calls EU Competition Commissioner, Margrethe Vestager, “the tax lady”. And the French plan to impose a three percent digital levy on the revenues of tech companies was countered by Mr Trump’s threat to impose swingeing tariffs on the import of French wines.


For all that, the weather is changing. For years, fund managers just had to own a basket of FANGs, sit back, and enjoy their outperformance. But it’s now been a year or more since these stocks have hit new highs. Facebook and Amazon are more than 10 percent down from their peaks; Netflix has lost about a third of its value since its June 2018 high; and Google is the outperformer by only being down single-digit percentages from its all-time highs last year.

Amazon released its Q3 figures yesterday evening. Revenues were $US70 billion, up 24 percent on Q3 2018. Amazon Web Services (AWS) revenue was up 35 percent over 12 months to nearly $US9 billion. But EPS was $US4.23 against analysts’ expectations of $US4.62 per share. In after-hours trading its shares were down by as much as 8.6 percent.

Meanwhile, back in low-Earth orbit…

Last week I described how a British start-up called Reaction Engines has designed a revolutionary new propulsion system which could power aircraft and rockets to hypersonic speeds in the upper atmosphere and near-space. On Tuesday (22 October) Aviation Week revealed that Reaction had successfully run a test at Mach 5 temperatures, validating for the first time the capability of the innovative heat exchanger design to operate in hypersonic flight conditions.

The breakthrough test brings Reaction’s goal nearer of using a lightweight heat exchanger to boost high-speed turbojets for supersonic and hypersonic aircraft (or missiles) as well as for developing the company’s Synergistic Air-Breathing Rocket Engine (SABRE). This is designed to provide low-cost, repeatable access to space.

The Mach 5 test took place in the second week of October at the company’s test facility at the Colorado Air and Space Port near Watkins (just outside Denver International Airport). This test comes seven months after the heat exchanger demonstrated the potential to operate in conditions equal to Mach 3.3. Heated air for the tests is generated by a General Electric J79 turbojet engine. In the Mach 5 test, the temperature was reduced from around 1,000 degrees Centigrade to roughly 100C in less than one 20th of a second.

As I wrote last week private equity house Elliott Advisers, Odey Asset Management and BAE Systems are all heavily invested in Reaction. Popping over to New York for lunch may soon be quite possible – though I doubt if Saint Greta will approve…

STOP PRESS: Virgin Galactic is to launch (note the verb) on the New York market next week. Now that will be interesting.

Victor.Hill@masterinvestor.co.uk

[i]See: https://www.newscientist.com/article/2220968-its-official-google-has-achieved-quantum-supremacy/?utm_medium=email&utm_source=dailynewsletter&utm_campaign=24102019

[ii]See Forbes, 23 October 2019, available at: https://www.forbes.com/sites/johnkoetsier/2019/10/23/ibm-googles-quantum-supremacy-is-150-million-percent-wrong/.