This is like saying that "machines are safe". It makes not sense, there all all kind of machines you can design. You should study a specific machine case by case.

Gene modification is enormous, and open to all kinds of unintended consequences. Just because we don't understand it yet it does not mean it is safe(we modify genes without really understanding what they really do with detail, we are unconscious ignorant).

I believe people should be able to buy GMO food if they what , but with traceability, if something wrong happens, and something wrong always happens with everything, to know it soon and do something about it.

But traceability is exactly what companies like Monsanto want to destroy. Look at Volkswagen, when something wrong happens, with traceability the lawsuits alone could bankrupt the company if people have children with leukemia or autistic and a scientific connection is proven.

-GMO because tons of Roundup® has been dumped on it ~Is Roundup® safe?

-GMO for marketing or other purposes but which could have features harmful to humans. ~How much, if any, scientific & statistically valid tests have been done?

-GMO for any reason but could have features which disrupt the current flora & fauna of their environment. ~Can a problem be observed before it is too late to correct for it?

-GMO which results in expansion and/or defense of monoculture. ~Are consumers ok with low quality food with minimal variety?

-GMO which is almost certainly owned by Monsanto ~Are we prepared to cede most of the benefits (and control; cf terminator seeds) of GMOs to a ruthless, aloof megacorp?

"we modify genes without really understanding what they really do with detail"

Except that with genetic engineering, we know precisely what they do. Ironically previous ways of changing genes are a lot more random.

No, we don't. At best, we know a few of the reliable consequences of that gene locus. We know nothing of the proliferation that gene may have in a wild population, or how our modification will help or hamper the organism to adapt to a changing environment.

"We know nothing of the proliferation that gene may have in a wild population, or how our modification will help or hamper the organism to adapt to a changing environment."

We don't know exactly, but we do know a lot more than nothing.

Ultimately, this. You can test for very, very small numbers, but you can't test for zero. This is why, for example, even though lead has no biological function, extremely small "acceptable levels" are specified in regulations rather than simply saying zero lead. You can prove that, for example, lead levels in the air are less than 0.15 µg/m^3, or that they are not. You cannot do that for zero (nonzero is easy to verify, but zero is not).

There is also, of course, the fact that no food is safe in all scenarios. Even pure water is toxic at high enough doses. For that reason, it makes no sense to hold GMO food to even higher standards than non-GMO food: equal standards certainly make sense, but not higher.

Normal food has the advantage that we already know it has no disastrous ecological consequences, by virtue of the fact that is has coexisted in its environmental niche for a while. The more food is altered, the less sure we become about that.

As Taleb has pointed out, our standards for danger should match the risk/reward involved. (E.g., we're ok with fast-tracking unproven cancer treatments because the alternative is death. Likewise, we have high standards for a common cold treatment because the alternative is inconvenience.) Unfortunately, we have no backup plan for the planet, so we must be extra careful with what we introduce to it.

To me, saying we don't need more stringent standards for GMOs is like saying we can definitely burn $X billion more oil reserves and not worry about global warming.

Except with genetic engineering, by being precise as to how and where the new genes are inserted, we are actually altering it less than with traditional breeding or mutagenics. Those types of breeding have far less testing than transgenics. We can potentially create new genes with those older methods that have never existed before and never know about it.

This is something that taleb was criticised for by biologists and geneticists (and basically his reply was "that they are idiots", yup, great argument).

"Normal food has the advantage that we already know it has no disastrous ecological consequences".

There is no such thing as normal food: Take the sweet potato. It has been altered transgenically using the agrobacterium twice in its history without human intervention. The agrobacterium is the same bacteria used when doing transgenic GMO. Basically nature beat us to it a long time ago when it comes to transgenics.

You can breed a potato the old fashion way to contain high level of solanine (which can be poisonous). Its not how the genes are altered, its what they do that is key.

It is also a way to enact intellectual property regimes over the world's biggest industrial sector. It is worth looking at how this has affected other parts of the economy.

I wouldn't mind GMO labels at all - hell, I'd buy primarily GMO-containing products on purpose, out of spite. But there's enough insanity going on about the labels we already have (e.g. E-numbers paranoia) that I don't think labeling GMOs will end well.