What could conceivably impede the exponential progress? Pandemics.
Everything we have attained until now, is going to assist us in taking the next big leap of humanity, colonisation of the planets, acceleration beyond our scope of purview. Within five years, industry experts plan to have an AI hardware device developed that could possess 10 billion neurons using memristors. The hardware AI has been all smashed together into one nanochip, which is capable of identifying and segregating analogue and digital signals, with integrated neural network. AI can also achieve hyper-performance by mimicking our brain’s design and being able to work in parallel with other neurosynaptic chips. If the development progresses at this rate, neural simulation technology will accomplish brain-like performance by 2025.
Let’s go back to the performance of AlphaZero developed by DeepMind — a simple AI, which was able to learn 3000 years of strategy, in just under a few months. If humans are already on the way to create hardware that could potentially achieve exascale and be able to fit in a box, the AIs will inevitably be able to do much more by 2030. Hardware AIs are already hundreds of times faster than a typical deep neural network. Therefore, these AIs would be able to operate in the unexplored and uncharted territories. If algorithms that rely on traditional microprocessors can learn 500,000 years of strategy and intelligence under just one year, how fast would hardware AI be able to learn? Would it be that difficult for the neural networks to start comprehending belongingness and eventually reach self-actualisation? It sounds like a dire warning for a dystopian future, where artificial intelligence will precipitate a statutory crisis by creating more intelligent beings in comparison to humans. But the denouement is purely dependent on our work, the homo sapiens, and whether we are building the right technology that would help “build” the future.
The implementation of artificial intelligence would undeniably assist us with rapidly advancing many other technologies and reduce the timeframe for calculations that would take thousands of years. But we should carefully consider practical and ethical outcomes for every single product that is made with AI. Everything we build with AI helps us to take the next step towards the AGI. Therefore, ethics and morality of the outcome of every AI-enabled product must be meticulously assessed. It is essential to create fully automated manufacturing plants with robot-assembly lines; however, the same concept cannot be used to replace every single human worker on the planet. We need substantial development in the field of BCI, to be able to compete with AGI and mitigate circumstantial challenges.
“A human B/CI system mediated by neuralnanorobotics could empower individuals with instantaneous access to all cumulative human knowledge available in the cloud and significantly improve human learning capacities and intelligence. Further, it might transition immersive virtual and augmented realities to unprecedented levels, allowing for more meaningful experiences and fuller/richer expression for, and between, users. These enhancements may assist humanity to adapt emergent artificial intelligence systems as human-augmentation technologies, facilitating the mitigation of new challenges to the human species.” — Martins et al. (2019)
One of the biggest questions at this point in time is whether AGI is vital for scientific progress. It’s yet to be answered. If an AI can already learn so much in a short period, given that the processing power of computation is exponentially accelerating, it wouldn’t be long until an AGI performs the necessary calculation to find a reason to exist and eventually decide to preserve itself for an extended period. It wouldn’t be difficult for a deep hardware neural network to feel the reason to differentiate themselves as a different ‘species’ in possession of higher intelligence.
Designer babies, DNA hard drives, synthetic biofuels and plant-based meat are all jaw-dropping outcomes of the science of genome editing. Every single one of these inventions are necessary to travel beyond the stars. But we need to save ourselves first — utilising a technology to help prevent imminent threat makes it plausible, and the society will find more reasons to believe in its future.
We are in desperate need of more companies like Sherlock Biosciences and the type of technology they are developing right now. CRISPR is so versatile that you could literally make red meat without killing a single animal, and also help prevent an imminent pandemic. All of that could happen by using the same technology but different techniques. This process of engineering biology has taken another step due to the outbreak of COVID-19. The outcome of Cepheid, Sherlock and Mammoth’s research will prove whether or not CRISPR will sustain the test of time and eventual immunity.
What could conceivably impede the exponential progress? Pandemics. During a crisis like this, leaders need to make the correct decision. Defining the correct ethical strategy is becoming blurrier these days. The leaders and policy makers are also at the core of making that decision, because at the end of the day, they are steering the society. Beyond the potential of a technology, is the power of the leaders to steer the countries towards the correct path. We cannot fully utilise the power of a disruptive idea unless we do the right thing with it.
It is evident that the coronavirus has affected the elderly more than any other age group (Liu et al., 2020). But that didn’t leave the younger and healthier group of people safe. Some of the people may be healthy, but many have ended up in the hospital and could still be carrying the virus. CRISPR could fast-track the process of detection and also help in creating vaccines for future pandemics. But it is our leader’s ethical responsibility to deal with this scenario. For instance, when the next pandemic hits, how are the governments going to react? When the anti-viral for COVID-19 is to be released, who is going to receive the first batch of treatment? Who will be prioritised? How long do people need to wait to receive the treatment? What will happen to the third-world countries? Governments around the world are talking about flattening the curve, but that would only be successful if they make the right decision before it all goes out of order. Besides, PHE in the UK discussed “herd immunity”, which simply didn’t work, and instead, and put millions of lives at risk.
Let’s say another pandemic similar to the novel coronavirus comes back within this decade. With a CFR rate of 5%, if 80% of the population in the UK alone is affected by that virus, then it will put approximately 2.6 million people at the risk of death. Does this approach mean that the herd immunity could be reached at the risk of 2.6 million people dying? While policymakers could divert all their resources towards NHS, e.g. increasing pays for physicians, or recruiting doctors, the approach of herd immunity does not make any sense given that all the scientific evidence points to the contrary. Before we reach for the stars, we need to save humanity. We need more investment in healthcare — the policymakers could start building emergency care centres and hospitals — recruit more staff for 111 and 999 medical emergency care. An outbreak could comeback anytime, therefore, the government should focus on gathering workforce, instead of piling up weapons for an imminent war. We must take lesson from this appalling pandemic that this was nothing but a war — not against humans — but the brutal side of mother nature.
We need to prepare ourselves for the next bio-threat, and not allow the human body to find a “natural course in due time”. Experts emphasised that tracing infections is important and that it’s mind-boggling that the leaders think about any alternative approach to this matter. Justifiably, it isn’t rational to put more pressure on the national health service. However, that leads to the data being skewed to a greater extent. If the authorities aren’t going to trace every single case, then the ratio between the actual number of patients infected versus the reported number of patients will likely be far lower, which will affect policy making. We cannot repeat this in the future.
This article is an excerpt from the book “Exponential Progress”.
Refs:
R. B. Martins, N., Angelica, N., Chakravarthy, K., Svidinenko, Y., J. Boehm, F., Opris, I., A. Lebedev, M., Swan, M. A. Garan, S., V. Rosenfeld, J. (2019) Human Brain/Cloud Interface. Frontiers in Neuroscience. 13:112, DOI: 10.3389/fnins.2019.00112.
Liu, K., Chen, Y., Lin, R. and Han, K. (2020). Clinical features of COVID-19 in elderly patients: A comparison with young and middle-aged patients. Journal of Infection, 80(6), 14-18.
