Skip to content

Comment on 'Artificial life' breakthrough announced by scientists

Comments

Call me a luddite, genetic engineering gives me the willies.

Luddite.

In all seriousness, the dangers brought up in the article can only come from a person lacking significant knowledge of bacteria.

Unimaginable numbers of bacteria permeate every part of the world, including miles within the earth, deep sea volcanic vents, radioactive waste, the rain drops falling on our heads, and us, up our noses, in our skin pores, our guts, etc.

Those simple bacteria evolve at staggering rates, and swap genes like DJ Danger Mouse mixing tracks.

Any new synthetic, or extraterrestrial, or magical, life form would have to compete. And unless you believe in magic it is unlikely to out-compete in a pre-existing niche.

The grey goo danger often thrown at nano-tech, is possible only if we violate one or more of the basic laws of thermodynamics.

And when it comes to antibiotic resistant bacteria, well we are already doing a great job of breeding those within ourselves.

But if you grew a barrel full of MRSA and then poured it down your toilet, the SA wold quickly lose its metabolically expensive MR. Because there is no Methicillin in the sewer, the resistance to Methicillin is a disadvantage. And the only way the Staphylococcus aureus could survive is if it dumped this feature.

But luckily you're probably taking antibiotics so you are a niche where MRSA can out-compete and thrive.

> And unless you believe in magic it is unlikely to out-compete in a pre-existing niche.

Cars are faster than land animals. Planes are faster than birds. Not being constrained by needing to have come from slow gradual iteration might open some possibilities, some terrifying. Biology doesn't produce the wheel.

To consider this as a possibility is not to believe in magic.

> Cars are faster than land animals. Planes are faster than birds.

There's no comparison.

I'm sitting here looking out the window at my car, and there are caterpillars out there that are moving faster than it. Not only that, but that caterpillar can refuel itself. And it'll turn itself into new ones next year. In the meantime, my car will sit and rust.

Really, it's the height of hubris to think that we can engineer something that works as well as what untold generations has evolved.

(And I'll apologize for the tangent, but this is also the fatal conceit of those who believe that any central authority can manage our economy any better than the market can on its own.)

"Really, it's the height of hubris to think that we can engineer something that works as well as what untold generations has evolved."

Why? Technological progress through logical deduction is patently more efficient than evolution through natural selection. Modern technology has already far surpassed the capabilities of natural organisms in a number of areas, all within a geologically minute timespan.

Just because a caterpillar can self-replicate, doesn't make it better than a car in every single respect, otherwise we'd be riding caterpillars to work. In terms of self-replication and regeneration, nature has us beat (for now). But if you want to send packets of information halfway around the globe, there is no biological process that comes anywhere close to our technological capability.

"(And I'll apologize for the tangent, but this is also the fatal conceit of those who believe that any central authority can manage our economy any better than the market can on its own.)"

But it can, or at least, a market with regulatory oversight is more efficient than a completely unregulated market place.

Natural environments and unregulated markets suffer from very similar problems. They don't magically gravitate toward the most efficient solution; very often they'll do incredibly stupid things that would leave an intelligent observer scratching their head.

However, natural selection does produce very robust ecosystems, which is why life has flourished for millions of years, and why caterpillars will outlast your car.

Logically, the best approach is to combine natural selection with intelligent oversight. Left to its own devices, an unregulated market has a tendency to fall into traps which benefits an individual or subgroup for a short time, but disadvantages the market in the long run. The Prisoner's Dilemma, the Tragedy of the Commons and the Market of Lemons are all common effects that can be circumvented by intelligent regulators, but would cause great damage to any market left to its own devices.

Modern technology has already far surpassed the capabilities of natural organisms in a number of areas, all within a geologically minute timespan.

They key words here are: "in a number of areas"

In a number of areas indeed. Can a plane fly without oil drillers, oil refiners, fuel trucks, and jet fuel?

Or in a more generalized form: For some X, thing Y is better at X.

Having said that, I in no way belong to those people who think that nature is magical and humans can never do better.

We can quite easily make incremental improvements some of the time.

Note that I consider it an improvement if it still does everything else equally well and some new things better.

If on the other hand, it can't do some things, then it is only useful in a number of areas.

So a plane is not an improvement on a bird, but a bird immune to some pesticide would be an improvement IF that immunity didn't have a huge cost.

Or we can say it's better only when lots of that pesticide is in its environment and worse otherwise.

I think quoting oil drillers is not really opportune at the moment if you want to give an example of how the damage from technology is going to be 'limited'.

See there's thing thing called an apple and this other thing called an orange.

And there's this discussion about artificial and natural micro organisms competing in the same niche. Like using light energy to capture CO2 for example.

And then there's oil drillers.

"Why? Technological progress through logical deduction is patently more efficient than evolution through natural selection. Modern technology has already far surpassed the capabilities of natural organisms in a number of areas, all within a geologically minute timespan."

There's a lot of hubris packed up into that statement. I really don't think you're considering the whole effort, the economy, the resources, the externalities that are involved in all those things that "surpass" natural organisms. The real place we've surpassed natural organisms is in many ways in global organization. Those things that we consider faster, stronger, better etc. than natural organisms are not self-standing things, but things that require our care and attention and vast amounts of energy, resources and infrastructure in upkeep.

And it's far from clear to me that logical deduction is more efficient than evolution. Evolution can certainly result in silly designs, that's for sure (the retina's layering, vas deferens' route, etc.); but evolution also has an appetite for complexity and resilience far exceeding anything the human mind can grasp today, aided or not.

"I really don't think you're considering the whole effort, the economy, the resources, the externalities that are involved in all those things that "surpass" natural organisms."

No, I have, and that's precisely the point. Our technology doesn't exist in a vacuum; it's an integral part of our civilisation. Machines do require an enormous amount of resources to back them, but so what? It's the end result that matters.

Many cells in nature also require relatively vast resources to be of use. For example, a mammal sperm cell requires billions of cells and massive amounts of resources to be of use. Take the sperm cell out of the body, and it will die far quicker than your car will rust.

"And it's far from clear to me that logical deduction is more efficient than evolution. Evolution can certainly result in silly designs, that's for sure (the retina's layering, vas deferens' route, etc.); but evolution also has an appetite for complexity and resilience far exceeding anything the human mind can grasp today, aided or not."

Yes, but it has had a few billion years head start, whilst the scientific method is only a few hundred years old.

From a computer science perspective, evolution and science both involve accumulating information and processing this information to produce useful designs. The pace of change, therefore, is dictated by how fast information can be gathered, and how much processing power can be applied to it.

Bacteria are particularly successful in this area. They are continuously passing around information and testing new designs. There is no intelligence involved in this process, but they make up for it through sheer numbers. Bacteria are, essentially, an unimaginably vast distributed computing network, continuously sifting through information and applying genetic algorithms to it.

We have a far smaller computing network, made up of our own neurons and our manufactured microchips. However, despite being trillions of times smaller, our networks are vastly more efficient. If they weren't, then death rates from bacterial infection would remain constant, yet they've been falling globally, and in developed countries, dying of a bacterial infection is today extremely rare.

But by aid of a car, you can travel faster than a cheetah.

You can make the argument that evolution does produce wheels, if only by the circuitous route of producing an organism smart enough to invent wheels.

The goal of engineering isn't to beat evolution; it is evolution at work, probing an as yet unexplored avenue of development.

> Really, it's the height of hubris to think that we can engineer something that works as well as what untold generations has evolved.

At best photosynthesis have an efficiency of about 8% At best solar panels have an efficiency of 40%

Is there a man-made process that can restore oxygen from carbon dioxide more effectively than photosynthesis? It also has to be reasonably cheap though ...

Yeah, but you can't eat solar panels.

> but this is also the fatal conceit

I think considering the possibility that removing gradual iteration as a constraint, or considering the big picture, could lead to improvements of some natural designs, true or not, is unlikely to be fatal!

I don't see how anyone would just dismiss this point a priori. Both of those things seem, intuitively, to be fundamental constraints in evolution. Human biology plus technology leads to us out-competing every other similarly sized animal in any land niche we choose. The question is whether bacteria biology plus human ingenuity could have any similar advantages over biology alone. If you trust your snap judgment without extremely careful consideration, I just hope you're not in charge of the relevant policy decisions!

> Really, it's the height of hubris to think that we can engineer something that works as well as what untold generations has evolved.

Even if true, it's irrelevant because nature's definition of "works" isn't always mine.

It isn't yours either.

For example, you're not using caterpillars or other "nature" for transportation so either you're stupid or you don't believe that caterpillars are superior to your car for certain purposes. (Caterpillars are cheaper, so you didn't choose the car for its price/performance either.)

Cars are faster than land animals. Planes are faster than birds.

Fuel, jet fuel.

Biology doesn't produce the wheel.

Tumble weeds? Insects which curl up into a wheel to quickly roll away from danger? Can a car drive up stairs?

Indeed if we genetically engineered a large animal to run on gasoline AND provided it with an ample supply of gasoline, it would move faster than anything else... as long as it had gasoline.

If we engineered a bacterium to eat oil and/or gasoline and dispersed in the gulf.... already there, eating oil, consuming the oxygen in the water in the process.

If we added it to an oil well... already there, lots of bacteria deep within the earth, eating everything from metal to methane. Probably not consuming the world's oil reserve only for lack of oxygen or any other oxidizer.

...some possibilities, some terrifying.

Ebola, HIV, the flu de jour all quite terrifying, all 100% natural.

And unlike you, they don't care. They will keep evolving and trying to kill you as fast and as hard as they can, every minute of every day.

> Ebola, HIV, the flu de jour all quite terrifying, all 100% natural.

If you can't imagine anything more terrifying than those, it is a failure of imagination. See Jared Diamond's point about how when humans and large animals coevolved gradually in Africa, the animals could gradually adapt, whereas when fully evolved homo sapiens were introduced in the rest of the world 40,000 years ago, almost all the megafauna were killed off almost immediately.

> They will keep evolving and trying to kill you as fast and as hard as they can,

Thankfully, this is wrong. Diseases that kill you quickly are at a severe disadvantage.

I can easily imagine an HIV that's as easy to transmit as the common cold.

Franky we got lucky with HIV, it could have been something like that and we along with cheetahs and a few other species don't have much genetic diversity within the species and are thus susceptible to extinction via disease.

This is why I want the biotech pedal to the metal. Because in my opinion the probably of us accidental creating something horrific are far less then something horrific being created quite naturally. Like in some shared living space with humans and domestic animals, it could happen in any given moment and spread quickly around the world. We don't need labs to die of something horrible.

> Indeed if we genetically engineered a large animal to run on gasoline AND provided it with an ample supply of gasoline, it would move faster than anything else... as long as it had gasoline.

Do you think gasoline is magic? Gasoline has almost exactly the same energy content as fat. Tons of animals have access to fat.

I think there's a risk that you're underestimating the scale required to make those things work. Not only are cars and their engines pretty big, but they need a huge amount of infrastructure for their construction (think about where that all that metal came from, never mind everything else), maintenance, feeding (the whole oil economy), roads, signals, piloting, etc.

It is inevitable that we will code a biological species that could not have arisen from gradual iteration.

However, there are some wheel-like things in biology: http://en.wikipedia.org/wiki/Rotating_locomotion_in_living_s...

This is a good point. At the same time, bacteria have been mutating for much longer than birds or land animals, and they iterate generationally at an unbelievable speed compared to their multi-cellular cousins.

The risks should be taken seriously, even the seemingly far-fetched ones, but need to be viewed in the context of an already cut-throat competitive landscape. IMHO, the bigger risk is intentionally designing malicious life forms (bio-weapons) more so than than beneficial ones getting out of hand.

It doesn't only bother me practically. Yes I did consider the fact that they'd have to compete, though I didn't consider the fact that they evolve all the time anyway, that's a good point. BUT, especially if they're talking about something as drastic as eating up a significant amount of carbon in the air, it will change the ecosystem. It'll pose the same danger as moving animals across continents.

But really it bothers me in a different way. I'll say morally, for a severe lack of better term. Maybe existentially. It doesn't sit right with me that we'll be living amongst creatures created in a lab, and that there's no way to differentiate them. I don't believe in God, but I like that we exist within a certain biological framework, it just feels like part of our identity. That the players are allowed to change the rules bothers me.

Your dog is genetically engineered. Your tomato. Your children. And I don't mean in the lab, but the usual way, by using your intelligence to choose suitable stock and then breeding until you get what you want. The shortcut in the lab is just faster and cheaper, not different in any basic or ethical way.

But that old-time genetic engineering still plays by the rules of nature, so to speak. Grafting branches between trees is pushing it a little. Splicing genes, not to mention synthesizing them from scratch, I think is pretty clearly of a different nature.

Golden delicious apples were a sport. Any really new feature in evolution is. The advantage of doing it in the lab is, its lots safer. You know what you're getting - just what you wanted. Not all the other random accidental changes you couldn't measure. Grandma's better tomato could actually make you sick - she had no idea what she was doing and no control over the process.

It has the potential to be different in the way that building a 747 is different than breeding a bird to be able to carry people across the Pacific Ocean. Either may be possible, but it's different.

I don't understand, you will never create a 747 with bird genes. In fact, they can do nothing with this technology that could not have been done by infinite monkeys with gene-splicing equipment. The only difference I see is the speed.

Luddite^2

>Those simple bacteria evolve at staggering rates, and swap genes like DJ Danger Mouse mixing tracks.

But they only evolve to the brink of failure

http://crnano.typepad.com/crnblog/2009/05/natural-protein-vs...

Life on earth has been trying to turn the planet into grey goo for billions of years already. I think we'll be OK.

I'm with you, but it is pretty fascinating all the same.

Yeah it's definitely cool.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.