) June 25, 2019 -- Tej Kohli, the London-based tech investor and entrepreneur, says that the hype around AI is still premature and that too much hype is creating a “bubble of enthusiasm” that risks causing disillusionment amongst the public when AI products do not start quickly improving their lives as promised. He says that AI will only be able to change the world when there is an accessible user interface which enables more entrepreneurs to build AI products that can enable all people to access the full benefits of the AI economy. This will require concerted international collaboration as well as major capital investment.
Kohli is ebullient about the prospects of AI to build a better world and has reiterated his belief that an exponential global AI economy could be worth $150 trillion by 2025; but also believes that it is premature to talk about the all-pervasive impact that AI will have on everyday life quite yet.
Kohli says: “The core problem with AI right now is that it does not yet have an accessible and intuitive user interface that allows any individual or company to ‘plug in’ to the AI economy. For AI to become manifest and omnipresent, it needs to be as easy to create AI products as it is to create smartphone apps. Currently it’s like we have all of the parts to build a car right in front of us, but it’s currently far too expensive to put those pieces together and create a car.”
Early technological progression often happens deceptively slowly. New technologies typically become disruptive only when entrepreneurs can intuitively access and use the technology to create new products and applications. The early Internet was almost useless until the first web browsers were launched by Netscape and Microsoft. Very early smartphones had limited functionality until Android and iOS became the dominant smartphone platforms upon which applications could easily be built. The first 3D printer was invented in 1984 but only started making a disruptive impact when contemporary 3D printers became available with an interface equivalent to a home printer that is intuitive, inexpensive and easy to use. Once that an intuitive and accessible interface is available for a technology, its growth often becomes exponential.
Every AI product currently relies on several elements that must work together. The Internet of Things is needed to connect sensors and other sources of data collection to enable learning. Big modalities of data then have to be stored in The Cloud and made available for access at high speed across 5G speed networks. Algorithms then need to be programmed to create probabilities and patterns and stochastics from collected data, and finally an output is required - most likely as some form of robotics - for AI to use its intelligence to take actions.
Whilst all of these things are currently accessible in ways that they never have been before, with cloud-based processors and the advent of 5G removing almost all cost barriers, and platforms like G2 enabling software developers to start integrating AI into their products and applications, making all of these ‘parts’ work together remains stubbornly expensive.
Kohli says: “The missing component is knowledge. We need more people working on solutions that dematerialise all of the separate components of AI into a user-friendly platform. Only then can it be democratised and used by entrepreneurs to build AI products. Whoever is able to do this first will have an unprecedented competitive advantage over all others.”
In 2018 the United Kingdom Government established the official ‘Office for Artificial Intelligence’ and announced £115m plans to fund 1,000 postgraduate students to study AI. But the UK still faces a massive skills shortage in the tech sector. According to the 2019 Morgan McKinley Salary Guide for Technology, demand means that permanent AI professionals can expect a 20.7% pay increase when switching positions.
In an attempt to create an accessible AI platform, IBM has made its Watson-branded AI services a free open source resource in The Cloud. But in February 2019 IBM had to start allowing enterprises to download Watson onto their own servers, in part to ameliorate the ongoing and expensive challenge of fusing many sources of Big Data into new AI products.
On 9 June United Technologies announced that its aerospace division will merge with defence contractor Raytheon so that the combined companies can “enjoy enhanced resources and financial flexibility to support significant R&D and capital investment”. By collaborating, United and Raytheon say that they will be able to work on a range of new advances in artificial intelligence. But the merger also highlights the sheer magnitude of capital investment that is needed to start making the visible components of AI work together to create AI products.
Data supports this view. A March 2019 survey of 2,830 ‘AI’ tech companies in 13 EU countries found that 40 per cent of those describing themselves as “AI start-ups” had no evidence of any machine-learning technologies that were “material” to what they actually did.
Kohli says: “AI could impact and improve lives many times over compared to the digital revolution. I prefer to call AI ‘alternate intelligence’ because it is not so artificial, it is much more precise and accurate and reliable than human intelligence. But to become omnipresent AI first needs to be demonetised, dematerialised and then democratised. If it becomes omnipresent across healthcare, biotech, hospitality, retail and thousands of other industries, AI could hugely improve everything about human life. But people will first need to be fully woken up to the vast possibilities for ‘alternate intelligence’ to significantly improve their lives. They will need to be willing to make the decision to let AI into their lives. But if their first experiences are overhyped, and then AI products do not start quickly improving their lives as promised, then the risk is that they will start to turn off AI before it even gets started.”
Kohli is an engineer by training who studied at the Indian Institute of Technology. Described as the equivalent of “Harvard, MIT and Princeton combined into one institution”, Kohli’s contemporaries at ITT comprise a ‘who’s who’ of technology pioneers and business leaders, including Vinod Khosla, who founded Sun Microsystems. Kohli himself has a long track record for spotting new trends and getting in early. He first made his fortune during the dotcom boom. When he saw the early potential for cities like Berlin to become new global hubs of technology, he invested heavily into real estate. When his son introduced him to esports, a hyper-growth global phenomenon, Kohli invested €20m into Europe’s leading esports team.
“We are so nearly there” says Kohli, “All of the pieces of AI are visible, and we know how they need to work together, but we haven’t quite solved the economics for allowing any individual or company to easily ‘plug in’ to the AI economy and develop their own AI products. Though there are a lot of people, including myself, who are investing in trying to get us there.”
As an example, Kohli points to Seldon, a London-based tech company that is backed in part by Rewired, an investment fund that counts Kohli amongst its lead investors. Seldon has developed the world’s first open source machine learning platform to enable data scientists, developers and large organisations to share data and harness the network effect to enable machines to learn faster, bigger and better. This will enable companies to create and implement AI products, and Kohli predicts it will propel Seldon into quickly becoming a huge business.
Kohli also points to Aromyx, a US-based company whose early technology was funded by DARPA, but which is now backed by Rewired. Aromyx is developing technologies that will enable the measurement and digitisation of taste and scent, which will allow for the detection and capture of new data modalities for the machine learning that is essential for AI. If successful, the US company will own the world’s platform for the capture and transfer of scent and taste data into AI products, with many enterprise, defence and consumer applications.
As for the more sensational aspect of artificial intelligence: the question of whether AI will one day subjugate the human race, Kohli is unequivocal: “AI is not about tech. AI is about making every industry and everything that is around us perform better, with minimal human input. Humans will always be in control of AI and will always have the ultimate say.”
About Tej Kohli
Tej Kohli is a London-based technology and real estate entrepreneur, businessman and philanthropist who now focuses on visionary ventures and investments that have the potential to transform lives and change the world. Through his investment vehicle Kohli Ventures, Mr Kohli targets high-impact investments into AI, robotics, biotech, genomics and Fintech that help to solve the world’s most pressing challenges and create a better tomorrow.
Mr Kohli also has a well-publicized mission to cure corneal blindness worldwide by 2030 through the philanthropic Tej Kohli Foundation that provides direct support for the Tej Kohli Cornea Institute, a preeminent global institute for solving corneal blindness worldwide.
Mr Kohli is a Distinguished Alumni of the Indian Institute of Technology in Kanpur (IITK) where he completed a degree in Electrical Engineering. He first rose to success during the dotcom boom selling technology solutions and e-commerce payments software.
Rewired is a robotics-focused venture studio specializing in applied science and technologies. Rewired launched in 2017 with an initial $100m fund for technologies that will enable robots to interact properly with real-world environments. In backing the development of robots’ sensory capabilities and understanding of the world around them Rewired seeks to play a major part in enabling the safe and efficient integration of AI into everyday life.
A video about Rewired can be accessed here: https://youtu.be/6yc5EcBIU5U
Photos/Multimedia Gallery Available: https://www.businesswire.com/news/home/52002582/en