Genetically Modified Organisms Risk Global Ruin, Says Black Swan Author

Experts have severely underestimated the risks of genetically modified food, says a group of researchers lead by Nassim Nicholas Taleb

It is 20 years since the FDA approved the Flavr Savr tomato for human consumption, the first genetically engineered food to gain this status. Since then, genetically modified food has become a significant part of the human diet in many parts of the world, particularly in the US. In 2013 roughly 85 per cent of corn and 90 per cent of soybeans produced in the US were genetically modified.

Given the ubiquity of this kind of foodstuff, you could be forgiven for thinking that the scientific debate over its safety has been largely settled. It is certainly true that a large number of scientists seem to take that view. In 2012, for example, the American Association for the Advancement of Science declared that genetically modified crops pose no greater risk than the same foods made from crops modified by conventional plant breeding techniques.

Today, Nassim Nicholas Taleb at New York University and a few pals say that this kind of thinking vastly underestimates the threat posed by genetically modified organisms. “Genetically modified organisms represent a public risk of global harm,” they say. Consequently, this risk should be treated differently from those that only have the potential for local harm. “The precautionary principle should be used to prescribe severe of limits on genetically modified organisms,” they conclude.

Taleb and co begin by making a clear distinction between risks with consequences that are local and those with consequences that have the potential to cause global ruin. When global harm is possible, an action must be avoided unless there is scientific near-certainty that it is safe. This approach is known as the precautionary principle.

The question, of course, is when the precautionary principle should be applied. Taleb and co begin by saying that their aim is to place the precautionary principle within a formal statistical structure that is grounded in probability theory and the properties of complex systems. “Our aim is to allow decision-makers to discern which circumstances require the use of the precautionary principle and in which cases evoking the precautionary principle is inappropriate.”

Their argument begins by dividing potential harm into two types.

• The first is localised and non-spreading. 
• The second is propagating harm that results in irreversible and widespread damage. 

Taleb and co say that traditional decision-making strategies focus on the first type of risk where the harm is localised and the risk is easy to calculate from past data.

In this case, it is always possible to make a mistake when decision-making about risk. The crucial point is that when the harm is localised, the potential danger from a miscalculation is bounded.

By contrast, harm that is able to propagate on a global scale is entirely different. “The possibility of irreversible and widespread damage raises different questions about the nature of decision-making and what risks can be reasonably taken,” say Taleb and co. In this case, the potential danger from a miscalculation can be essentially infinite. It is in this category of total ruin problems that the precautionary principle comes into play, they say.

A key difference between these types of risks is the statistical structure of their impact. This structure is either dependent on scale or independent of scale. An example of a scale -dependent distribution is the weight of an adult human which is generally never greater than about 10 times the weight of an average human.

However, a single individual can be richer than the poorest 2 billion humans. So wealth follows a scale independent distribution.

When it comes to risk, the harm from scale dependent risks comes from the collective effect of many events, since no single event alone can dominate the total. “It is practically impossible for a single data account for 99 percent of all heart attacks in a given year,” say Taleb and co by way of an example.

By contrast, the harm from scale independent risks can be dominated by a single event. For example, the Earth is constantly bombarded by small rocks from space that have little effect. Nevertheless, a single large rock could wipe out the entire human race.

Taleb and co focus on two examples. The first is nuclear energy. They point out that many people are justifiably concerned about the risk associated with nuclear energy. Scientists are well aware of the harm that can be caused by radiation release, core meltdown and the disposal of radioactive waste and these risks have been studied extensively.

While the potential harm from a nuclear accident can be large, it is generally scale dependent and far from global. So when it comes to making decisions about whether to use nuclear energy on a local scale, the risks involved can be managed using appropriate safety measures that have been carefully considered.

Taleb and co contrast this to the case of genetically modified organisms. They argue that the risk from genetically modified organisms is a potential for widespread impact on the ecosystem and widespread impact on human health. In other words, it is scale independent.

One of the arguments that genetically modified crops are safe is that it is no more unnatural than the selective farming that people have been doing for generations. However, Taleb and co argue that this kind of farming is different from the current practice because any mistake in the form of a harmful variation will almost certainly be localised and die out as a result. This is the natural process of selection.

Over many generations, humans have chosen and adapted biological organisms that are relatively safe for consumption, even though there are many organisms that are not safe, including parts of and varieties of the crops that we do cultivate.

By contrast, genetic engineering works in a very different way. This process introduces rapid changes on a global scale. But selection cannot operate on this scale, they argue.

“There is no comparison between tinkering with the selective breeding of genetic components of organisms that have previously undergone extensive histories of selection and the top-down engineering of taking a gene from a fish and putting it into a tomato,” they argue. “Saying that such a product is natural misses the process of natural selection by which things become “natural.””

The potential impact of genetically modified organisms on human health is even more worrying. Taleb and co say that the current mechanism for determining whether or not the genetic engineering of particular protein into a plant is safe is woefully inadequate.

The FDA currently does this by considering the existing knowledge of risks associated with that protein. “The number of ways such an evaluation can be an error is large,” they say.

That’s because proteins in living organisms are part of complex chemical networks. In general, the effect of a new protein on this network is difficult to predict even though the purpose of introducing it is to strongly impact the chemical functions of the plant, for example, by modifying its resistance to other chemicals such as herbicides or pesticides.

Even more serious is the introduction of monocultures— the use of single crops over large areas. This dramatically increases the likelihood that the entire crop might fail due to the action of some invasive species, disease or change in the environment.

When harm is localised, it can be used as part of the learning process to prevent the same set of circumstances occurring again. Global harm is different. “We should exert the precautionary principle here because we do not want to discover errors after considerable and irreversible environmental and health damage,” conclude Taleb and co.

They go on to discuss a number of fallacious arguments against using the precautionary principle. One of these is the Loch Ness fallacy, which states that the precautionary principle should prevent us from swimming in the Loch because we have no evidence that the Loch Ness monster does not exist.

Taleb and co say this is a corruption of the absence of evidence problem and unrelated to the question of whether the precautionary principle should apply. That is because the harm associated with the Loch Ness monster is purely local. “If the Loch Ness monster did exist, it would still be no reason to invoke the precautionary principle, as the harm he might cause is limited in scope to Loch Ness itself, and does not present the risk of ruin,” they say.

They go on to consider numerous other fallacies that confuse the issue over whether to use the precautionary principle or not. The central point in most of these is whether the risk involved is one of global ruin or local ruin.

That is an interesting contribution to the debate over genetically modified organisms, which has become becalmed in recent years. While the argument itself is interesting, the fact that the lead author, Nassim Nicholas Taleb is such a high profile commentator on risk is bound to raise the profile of the debate. The co-authors include a number of other well-known researchers such as Raphael Douady at the Institute of Mathematics and Theoretical Physics in Paris and Yaneer Bar-Yam at the New England Complex Systems Institute in Cambridge.

It is unlikely, of course, that the other actors in this debate will meekly agree with this new assessment. In particular, companies with a financial interest in the future of genetic engineering, such as Monsanto, will want to put their side of the debate.

More interesting will be how the FDA takes into account the arguments that Taleb and co put forward. Expect fireworks in the coming weeks and months.

Ref: arxiv.org/abs/1410.5787 : The Precautionary Principle (with Application to the Genetic Modification of Organisms)

The Physics arXiv Blog

ORIGINAL: Medium

by The Physics arXiv Blog