By Professor Louis C H Fourie
JOHANNESBURG - In 2019 Amy Webb, the chief executive of the Future Today Institute in the US, wrote a very interesting book called “The big nine: How the Tech Titans and their thinking machines could warp humanity” in which she discusses the most pressing problems around Artificial Intellagence (AI), possible scenarios of what could happen in the future and a solution of what could be done to ensure that AI is on the right path.
The “Big Nine” tech companies
The “Big Nine” is a reference to the nine biggest technology companies who plays a dominant role with regard to AI.
Six of them, collectively referred to by Webb as the G-MAFIA, are in the US: Google, Microsoft, Apple, Facebook, IBM, and Amazon.
She refers to these six companies as “a closed supernetwork” that exerts a large influence on our lives and to some extent even controls our futures.
The three remaining companies of the “Big Nine” are the Chinese technology heavyweights Baidu, Alibaba, and Tencent, who are often referred to as the BAT companies.
Alibaba, founded by Jack Ma and friends in 1999, is the world’s largest retailer and e-commerce company, and the fifth largest AI company. Tencent was founded in 1998 and provides various Internet-related services and products, including AI, entertainment, games, and other technologies.
Tencent is well-known in South Africa, since Naspers owns just more than 30 percent of this R6.724 trillion company with the 5th highest brand value in the world and stakes in more than 600 companies worldwide. Baidu has a market value of R677.7 billion and specialises in Internet-related services and products, as well as AI.
The frightening power of the Big Nine
Together these nine large technology companies control most of the AI world. This is very significant, since AI has been integrated into so many devices, services and decisions that people are increasingly experiencing the power of algorithms in dynamic pricing when buying online, applying online for a loan, or even applying for a job. And we are often powerless since you cannot argue with or appeal to an algorithm.
People are, therefore, rightfully concerned about the future of the human race and typically blame the large tech companies. And indeed, these large companies do wield a scary amount of influence through their control of operating systems and thousands of applications on billions of personal, household and office devices. They are responsible for many of the current day concerns regarding AI, whether directly or indirectly. They largely contributed to the problems of algorithmic bias (that I discussed on 9 Oct. in this column), the distorted reality of social media filter bubbles and echo chambers (discussed on 16 Oct.), lack of diversity, and numerous privacy issues.
This does not mean that the Big Nine are necessarily and intrinsically evil. They do see the potential of AI to improve health care and longevity, to streamline the work environment, as well as to solve our impending climate and other complex challenges. The problems lie much deeper in the underlying systems and the external forces discreetly putting pressure on them to operate in a certain way.
Commercialisation of AI research
Over the years, AI research has gone through the typical hype cycle – periods of extreme excitement, unrealistic hype, and an abundance of money, followed by periods of absolute disenchantment and a lack of AI research funding when the technologies did not deliver on the over-optimistic promises and tight timelines.
Research on AI, especially the currently popular neural networks and machine learning, requires huge amounts of data, computer power, and therefore a significant investment in money. Due to dwindling support from governments in the higher education environment, many AI researchers and scientists turned to the large technology companies for funding.
Although many AI projects were saved by the deep pockets of the large technology companies, it came at a price since these companies are profit driven and need to show a return on every investment. Since the digital world is a very competitive space, where quick wins and windfalls are the norm, the G-MAFIA and BAT companies placed tremendous pressure on the research laboratories and researchers to build practical and commercial applications for AI as fast as possible.
This not only led to the release of premature AI products, but more importantly the commercialisation of AI research. This commercialisation is clearly seen by the large number of research labs and start-ups that have been acquired by large tech companies, such as Google that bought Nest Labs for R528 billion and DeepMind for R8.3bn in 2014; Apple that acquired Turi Inc for R3.3bnin 2016; Intel that bought Nervana for R5.8bn and Movidius for R6.6bn in 2016; and the very expensive acquisition of CTRL-Labs by Facebook for a massive R16.5 billion in 2019, to name but a few. Others are funded by large companies as in the case of OpenAI that is funded to the tune of R16.5bn by Microsoft, Elon Musk (SpaceX, Tesla, and Neuralink) and Sam Altman previously from Y Combinator and currently the chief executive of Open AI.
Disappearance of small AI companies
However, it is a serious concern that the majority of promising AI start-ups have been bought up by the large tech companies, frequently ending public availability of any AI products they created. The top acquirers have been Apple, followed by Google, Microsoft, Facebook and Amazon. In 2019 a record number of 231 acquisitions of AI start-ups have been made, up from 42 in 2014.
This enormous number of acquisitions is disquieting, since AI is perhaps one of the most important technologies of our lifetime and is informing or making an increasing number of decisions and unfortunately has been susceptible to algorithmic bias, the creation of filter bubbles, unethical use such as facial recognition in China, and irresponsible military use. It is, therefore, particularly important to have a broad, diverse research and business community developing AI solutions.
Locked into a controlled ecosystem
And as the G-MAFIA and BAT companies are wielding more power and the AI technology they created permeate almost every aspect of our lives, the situation is growing increasingly precarious. The large tech companies are becoming the operating systems and decision controllers for everyday life for billions of people. Once your smart devices, phones, gadgets, appliances, and cars are connected, the gathering of data points and profiling starts, and you are locked into a carefully controlled ecosystem.
In total contrast to the G-MAFIA companies, all AI companies in China are fully controlled by the state. The BAT companies are legally bound to put their technology and data at the disposal of the state to enable them to conduct mass surveillance of citizens. This takes the control over people to a totally new level.
The infamous Sesame social credit system of China uses AI algorithms running on the platforms of the BAT companies to keep a close watch on the behaviour of Chinese citizens. Good behaviour and the following of rules are rewarded, while bad behaviour such as jaywalking or playing video games late into the night, are punished.
The need for an AI framework
AI is not necessarily detrimental and has brought tremendous benefits to everyday life. But the G-MAFIA companies are profit-driven and need to satisfy their shareholders, while in China the BAT companies must please the Chinese government. The problem is, however, that what is best for shareholders and the Chinese government, may not be in the best interest of humanity.
Adversarial attacks, algorithmic discrimination and surveillance control will continue to harm unsuspecting users across the world. The social, economic and political divide continues to grow, while people have lost ownership of their personal data, their privacy, and their identities.
Unless, of course, countries enforce certain AI frameworks, standards, and best practices to ensure that AI is used for the human good. AI practices should be transparent with standardised protocols, and citizens should have control over their data. If we do not fix the problems at the root, the consequences in the future can be disastrous.
Professor Louis C H Fourie is a futurist and technology strategist.
BUSINESS REPORT