Enhancing lessons learnt and redistributing them to the entire ecosystem is a cornerstone of safety enhancement. It is much facilitated in the case of Artificial Intelligence (AI) thanks to the remote update possibility, as demonstrated successfully by Tesla.
Implementing a statistical approach instead of a deterministic one. Some statistical risk analysis approaches are already available for years in the form of fault trees to determine the statistical probability of a feared accident. However this only works in environments where statistical failure data of components is available, and with limited changes to the environment and the system. New statistical approaches will have to be developed based on specific testing of the entire AI-related system. These approaches need to be developed theoretically and empirically and remain the major challenge of the years to come.
Rules governing operability of the system in case of component failure will have to be strictly defined and enforced (with how many sensors out of order is it safe to drive autonomously?), because the degraded situations are the most difficult and cumbersome to regulate.
The problem of the new statistical approaches to safety demonstration is an exciting problem facing all regulators. I am looking for some science behind this, if any reader has useful links please share!
Our current approaches to the regulation of system risk management and prevention of deadly accidents remains very much deterministic. In the most critical applications such as in nuclear power plants or aircraft controls, regulatory authorities require a deterministic demonstration of the links between input and outputs. Superfluous code that is not used needs to be removed, just in case. Older processors are used which reactions are fully known.
With the advances of Artificial Intelligence, this won’t be possible any more. In particular because the devices become black boxes that have learned to behave in a certain manner most of the time when exposed to certain stimulus. However deterministic proof of the relationship between input and output is impossible and we don’t quite know how it really works inside. It can only be a statistical measure.
This situation is an extensive challenge for the regulatory authorities that will have to regulate safety-critical applications based on AI such as automatic driving. Most current regulatory approaches will become obsolete.
Some regulatory authorities have identified this challenge but most have not, although this will constitute a real revolution in regulation.
While the development of drones and robots that could take themselves the decision to engage targets becomes closer, the issue of whether to develop such system becomes a conundrum. It is important to be able to face such a possible threat, at the same time usage of this type of weapon will need to remain very much controlled. Mechanisms similar to control of nuclear proliferation or chemical weapons might need to be put in place – with the particular challenge that no huge and noticeable industrial complex will be needed to produce such weapons.
The Open Letter by concerned scientists on autonomous weapons is interesting to read. It states “If any major military power pushes ahead with AI weapon development, a global arms race is virtually inevitable, and the endpoint of this technological trajectory is obvious: autonomous weapons will become the Kalashnikovs of tomorrow.”
At the same time, military might contend with enhancing human capabilities by teaming humans with robots, in particular to be able to take decisions in uncertain situations. But the issue needs to be tackled quickly because the consequences of robots engaging without control could become a proliferation issue.
One area that has struck me in particular is the reference to exceptional performance of the combination of human and artificial intelligence when is comes to pattern recognition. An example: “In one recent study, given images of lymph node cells, and asked to determine whether or not the cells contained cancer, an AI-based approach had a 7.5 percent error rate, where a human pathologist had a 3.5 percent error rate; a combined approach, using both AI and human input, lowered the error rate to 0.5 percent, representing an 85 percent reduction in error.”
It seems increasingly that this combination of two different perspectives, ours and the other one we are creating using Artificial Intelligence, could open us new frontiers. Artificial Intelligence in that sense seems rather a useful tool to broaden our perspectives and capabilities.
In the banking industry, it is estimated that 80% of the client value (i.e. fees) is still generated by the 5% face-to-face contact. The 95% client contact through internet and mobile does not generate much value. In a context of much lower returns in general for the financial industries, banks are confronted to a key dilemma: increase e-banking and convenience but without losing opportunity for creating revenue!
This is a typical example of the impact of the Fourth Revolution on institutions. Bringing services online is not just a transpose of the actual brick and mortar relationship and value chain. It creates the question of creating a whole new value proposition.
And it so happens as well that if simple transactions can easily be carried over to online interfaces, more complex transactions still require a more in-depth contact, either by phone or face-to-face. In the banking industry those transactions carry the most fees: investments and loans. But keeping branches open create significant fixed costs that see their return diminish. The manner of implementing those interactions in an online world still remains to be invented.
The main lesson for the moment is that by bringing current services and transactions online, believing that the value proposition will remain similar is an illusion. It will change significantly and it will need to be reinvented.
“Kill that bulky IT department!”. That could be the war cry of many organizations these days as the influence and size of IT departments tend to diminish significantly. As a result, they don’t have the same regulation impact on investment in Information Technologies.
Two related changes are driving this transformation:
the move to the Cloud (and thus the lesser need for infrastructure setup and maintenance), and
the related fact that other departments can now spend directly for systems without any infrastructure infrastructure needs and thus without any prior authorization or even knowledge by the IT department
According to Gartner, 38% of IT spend in companies is now out of the hands of the IT departments and this tends to increase significantly over time [reference: attended speech from CapGemini CEO in Oct 2016]. The marketing department in particular for BtoC industries, becomes a major client for information services.
This decentralisation has many positives. In particular it removes the centralizing controlling power of the CIO which was oftentimes excessive, even taking strategic decisions without proper understanding of the business impact. It allows specialist trades to implement the tools that they really require. On the other hand it opens the door to issues related to data consistency, possibilities of business intelligence, and all sorts of security-related issues for company data. Actual control of the expenditure may also become an issue as more and more cloud services are Opex based instead of being visible, centrally authorized Capex.
In any case, for us involved in providing specialist software (cf my company ProjectAppServices), it certainly means that all our marketing effort should be directly with the user, and the IT department is just an annoyance to avoid as much as possible.
Are you fully aware of this change? If your IT department still decides everything you are going into the wall. Time to change!
There are basically 2 types of resistance to social and technical change: the ‘luddite resistance’ and the ‘protected market’ resistance. And in the case of the Fourth Revolution, both add up when it comes to resisting change.
The ‘Luddite resistance’ refers to the Luddite, those early 19th century frame breakers, qualified weaver artisans that broke the new weaving machines that only required unqualified labor. Those workers feared for the end of their trade, being replaced by machines.
The ‘protected market’ resistance refers to all those markets that for one reason or the other, have been protected (generally in the name of the common good to manage scarce resources or to enhance certain types of services) but the reason for this protection disappears with the Fourth Revolution. In general, many scarce services and goods are now becoming abundant.
As an example, taxi services generally form a protected market with regulated prices and numbers, in the name of public service. Drivers fear to be replaced by machines (the luddite resistance) and at the same time, the reason for market protection disappears as ride-booking apps replace taxi hailing.
This resistance to change analysis framework is actually quite useful to analyse all sorts of resistance to change even in organizations today. People fear that their job will be replaced by machines or information systems, and at the same time the reason for the protection and regulation of certain trades disappears.
In your case, what is the proportion of ‘luddite resistance’ and of ‘protected market resistance’?
The premise is that the intrinsic complexity and sophistication of the empire or organization increases over time up to a point where additional complexity is detrimental, in particular in the face of sudden external change. The institution is then unable to cope with the change. “When societies fail to respond to reduced circumstances through orderly downsizing, it isn’t because they don’t want to, it’s because they can’t.”
I find this model intriguing because from my perspective, complexity rather increases reactivity and adaptation. I think the author mistaken complication and complexity. Adding layers of bureaucracy in a futile attempt at control is complication. Properly maintained complexity is rather an antidote at inflexibility. We should certainly fight organizational complication (and its representative, bureaucracy) but rather welcome complexity.
The paper is a bit long but worth reading: how a child suffering from schizophrenia found solace in his relationship with an African Shaman, discovering how his condition brought him closer to this special role. To read in Nautilus, ‘A Mental Disease by Any Other Name‘.
“Both shamans and schizophrenic people believe they have magical abilities, hear voices, and have out-of-body experiences.”
I find this document exciting because it shows that conditions that we judge as debilitating and requiring treatment in our society may have been rather considered a strange but valuable gift in other societies.
The document also shows how belonging to a community can help control symptoms associated with these conditions.
Let me stretch this observation to our framework of the Fourth Revolution. Our usual categorization of mental illnesses stems from the Industrial Revolution. What if it would be significantly upset by the Fourth Revolution. Moving into the Collaborative Age we might find that some of these conditions are gifts in certain situations too.
Becoming a contractor is an increasing trend: “[Independent contractors] share of total employment is rising, from 9% to almost 16% between 2005 and 2015. And it’s not just low-skill, uber-drivers turning to contract work out of desperation—the increase in alternative work spans all education levels. Americans with a college degree are most likely to be contract workers, and this group saw the biggest gains. Contingent work has also become more common across a variety of industries and occupations.”
One of the main issues with the fact that we will become increasingly contractors is to manage the risk of a sudden loss of revenue; and more generally, the ups and downs of income depending on how often we provide our services. This is a problem I am managing in my consulting company, voluntarily keeping a substantial share of earnings in the company to cope with periods with lower utilization. De facto, the company is being used as an income insurance buffer. It might not be the most efficient way, but it works.
The Quartz post proposes that the state could setup a ‘wage insurance’ against substantial drop of income to cover those extreme events that can really derail one’s life. This could be a very useful institution for the Collaborative Age, together with some sort of collective health and life insurance.
What other institutions could we think of for the Collaborative Age?
In the Industrial Age, job title was very much one’s social identity, in particular related to the position in pyramidal organization charts. In many countries like France, the studies (university, degree) and grade achievements was also very much one’s identity. It is still the case at various levels.
However, this easy-to-relate identify definition will disappear in the Collaborative Age as the importance of conventional organizations will progressively disappear, and as we will be increasingly on our own without a fixed ‘job’, or at least only with temporary ones.
This situation creates a lot of stress on personal identity. It is thus a high barrier for those that hesitate to jump out of traditional organizations; or, those who get retrenched or lose their job and have to reinvent themselves. It is possibly one of the biggest stressors in society today.
One needs to realize how defining oneself in terms of job title and university degree is limiting. In particular after a few years’ experience, our personal identity is much more complex and full; and it involves both personal and professional elements. We need definitely to find other ways of expressing our complete identity. It could be through our own creations or on social media.
Transforming the way we express our identity is a mandatory skill for the Collaborative Age.
It is soon the end of driver services such as taxi and limos, and those currently employed there should start looking for other occupations.
The self-driving car is just at the corner, and when society will realize that they are indeed much safer than human-driven cars, there will certainly be suddenly a tipping point.
Of course that will be felt like an unjust revolution by those employed in the driving trade, but let’s face it, that is clearly the direction of the world. Convenience, safety and efficiency will create the change. And this will impact also insurance companies, car manufacturing companies (because of the impact on car ownership) and a lot of related services (including the profitable industry of traffic speeding fines!).
I may even live to see human driving forbidden by insurance on public roads because of its dangerous nature.
It will be a disaster for those involved in the trade, and generally a great progress. They’s better anticipate it.