Fast-changing technologies: risks and benefits

Let’s start with some good news. Technologies have provided many huge benefits to humanity:

• Better health, via better medical treatments, better hygiene, and better nutrition
• More knowledge created and shared
• Greater availability of labour-saving devices
• Ability to travel more widely and have richer experiences.

However, it’s also the case that technologies are prone to have side-effects that are unforeseen and unwelcome:

• Drugs to treat medical conditions can have side-effects that are worse than the original sickness – an example being thalidomide, which caused horrific birth defects
• Widespread use of antibiotic drugs can stimulate the emergence of bacteria that are resistant to these drugs, threatening the spread of an untreatable infection
• Chemicals intended to control insects that carry diseases can damage other parts of the environment – an example being the insecticide DDT
• Chemicals added into industrial processes to boost performance or increase strength can have unexpected detrimental effects on the health of people in the vicinity – examples being lead (added to petrol) and asbestos (used in buildings)
• CFC chemicals used in refrigeration and in aerosols migrated high into the stratosphere, where they destabilised ozone molecules, allowing larger quantities of harmful UV radiation from the sun to reach the earth’s surface, increasing incidents of skin cancer
• A hydroelectric dam can generate significant quantities of electricity, but if the dam wall breaks, enormous damage can result to the people and habitat in the path of the water that rushes out; similar concerns apply to nuclear power stations with inadequate safety measures
• Innovative financial assets, such as Credit Default Swaps, Collateralised Debt Obligations, and other so-called credit derivatives, that can enable more investment in particular areas, can also destabilise financial markets, earning themselves the disparaging description “financial weapons of mass destruction”

That brings us to the 64 trillion dollar question:

How can the balance of risks and benefits be moved away from risks and toward greater benefits?

Technology risk factors

Other things being equal, a piece of technology is liable to cause more damage:

• If it is deployed at larger scale – for example, a larger hydroelectric dam may cause greater damage, if the dam wall breaks, than a smaller dam
• If it is deployed more widely – a chemical used throughout the world may have more drastic side-effects than a chemical used in just one location
• If it embodies more power – for example, high voltage electricity lines that are blown out of their intended routing may result in a more damaging discharge than low voltage lines
• If their method of operation is obscure and poorly understood – increasing the chance that they will operate in unexpected ways in subtly changed circumstances
• If the people or systems overseeing these technologies pay less attention, or lack adequate training or incentives to perform their tasks well
• If the people or systems overseeing these technologies cover up or downplay evidence of problems with the technology, and harass or ridicule people who want to highlight such evidence.

The above conditions can be called “technology risk factors”.

Unfortunately, several of these risk factors are likely to be increased when a new technology shows a lot of promise, and its backers are keen to take fuller advantage of it. In such a case, the backers would prefer:

• To deploy the technology at larger scale, and more widely, so that it can reach more people
• To run the technology at higher power, so that it is more likely to achieve striking results
• To minimise attention to any evidence of apparent behaviour that could cause the technology to be withdrawn.

Other risk factors will increase if the technology is complex and fast-changing:

• Its operating principles are less likely to be understood
• The people overseeing the technology are less likely to be adequately trained in all aspects of its operation.

There’s an additional set of complications when technology is changing quickly:

• An assessment of risks associated with the technology could become invalidated, due to the greater capabilities acquired in a subsequent version of the technology
• If the first group to develop that technology could gain a significant advantage, commercially or geopolitically, there will be incentives to cut corners with the implementation of safety measures, in a rush to obtain “first mover advantage”
• Fast-changing technology can alter, not only the primary solution being developed, but also the environment in which that solution operates, making reliable foresight harder.

Prioritising benefits?

One approach to the risk-benefit balance is to focus more on the benefits than the risks:

• Rather than slowing down the development and deployment of technology, from fear of damage arising, the priority would be to accelerate that development and deployment, in order to achieve the envisioned benefits more quickly
• Any unexpected and unwelcome side-effects can, in this approach, be addressed as and when they arise
• Responses to these side-effects will in any case be a valuable part of learning; any apparent “technology failures” can assist in faster acquisition of knowledge (about what works and what doesn’t work).

This approach – which is sometimes described as “techno-optimistic” – might be seen as especially attractive if the potential benefits from the new technology have huge scale:

• Large numbers of lives could be saved, by innovative new treatments for cancer, dementia, and so on
• New nuclear power plants could generate enormous quantities of green energy, allowing faster reduction of emissions of greenhouse gases
• Geoengineering interventions in the stratosphere – or by changing the composition of the oceans – could reverse the dangerous trend toward excessively high global temperatures
• AI algorithms with greater general intelligence could propose profound novel solutions to longstanding issues of healthcare, global warming, and the management of the financial and economic markets.

Techno-optimists urge, accordingly, a speed-up in the development and deployment of breakthrough innovative technologies.

The problem, nevertheless, is that any technology with greater potential to deliver profound benefits has, at the same time, greater potential for disastrous unexpected side-effects. Recall, again, the metaphor of the two-edged sword.

By itself, this is no reason to try to slow down the development and deployment of new technology. However, it is a reason to ramp up parallel efforts to anticipate and manage any side-effects that may arise.

Such efforts can be guided by the principles described in the pages ahead.

What about ethics?

The development and deployment of technologies is sometimes addressed from the viewpoint of ethics. Thus it is common to hear advocacy for “the ethical use of technology”, or, more simply, for “AI ethics”.

This framing can cause problems. Self-proclaimed ethicists are sometimes perceived as dour, unimaginative people who mainly say “no”, as in “thou shalt not do this” and “thou shalt not do that”. Understandably, technologists and entrepreneurs sometimes bristle, when they perceive that they are being upbraided by critics whose ideas appear to derive from contentious philosophical or theological worldviews. The technologists and entrepreneurs want to proceed with the task, as they see it, of building a significantly better world. They’re disinterested in interference from apparent do-gooders.

Moreover, the common ethical injunctions to ensure “fair” or “equitable” access to technologies are subject to controversy as well, since there are divergent views on what counts as “fair” and “equitable”. Just banging the table and saying “this needs to be more fair” is only the very beginning of a complex discussion.

Nevertheless, the Singularity Principles do endorse one of the fundamental insights of ethics:

• Just because we believe we could develop some technology, and even if we feel some desires to develop that technology, that’s not a sufficient reason for us actually to go ahead and develop it and deploy it.

More briefly: could does not imply should.

However, the principles take as their starting point an appeal, not to any elaborate ethical framework, but to considerations of potential harm and potential benefit – especially considerations of potential catastrophic harm or potential profound benefit.

It’s far from easy to calculate in advance the likely impacts of various technology projects on catastrophic harm or profound benefit. That’s where the guidelines of the Singularity Principles come to the fore. They provide a number of short-cut recommendations of the form, “It’s generally good to do this” or “It’s generally bad to do that”. Whether you agree with these recommendations may depend on your assessment of the extent that, on the whole, they will indeed alter the likelihood of catastrophic harm or profound benefit.

The transhumanist stance

I’ll insert here one additional quick comment on the subject of ethics. Answers to the question “what should we do” generally presuppose a view about which kinds of state of being are more virtuous.

One family of ethical views tends to regard the lives of certain people in the long-distant past as somehow the most virtuous. These people – such as founders and leading figures of religions – are said to provide role models for us to seek to emulate.

Another family of ethical views tends to look to elements of the present day, or the recent past, as models that the lifestyles of all people should attain.

The Singularity Principles make neither of these assumptions. Instead, they conform to what can be called “the transhumanist stance” – the view that people in the future (including the near future) can attain levels of wellbeing, consciousness, and virtue, that radically surpass what has been attained at any previous period in history.

Transhumanists do not accept that our vision should be constrained by the accomplishments and achievements of the past. Transhumanists anticipate a dramatic uplift in human capability – an uplift in capability that will be available to everyone.

What will enable that uplift is wise application of the possibilities provided by new technology – technologies that can significantly improve:

• Health, longevity, and resilience
• Intelligence, insight, and awareness
• Collaboration and creativity
• Sustainable living
• Mental and emotional wellbeing.

In short, the transhumanist stance is that it is possible, and desirable, to significantly improve all aspects of human life, by the wise application of science and technology, guided by clear thinking.

From this point of view, it would be doubly tragic if emerging fast-changing technologies were mismanaged. The first tragedy is the civilisational collapse that is likely to ensue. The second tragedy is humanity being prevented from reaching the much higher levels of flourishing that were so close to being attained.

Moving forward

The technology risk factors discussed above – and the range of benefits that technology can enable – apply to a wide spread of different technologies. Next, let’s consider how this picture changes when we focus on one particular technology, namely artificial intelligence.