Even with proper use, you can't avoid herbicide resistance. Selection pressure is selection pressure. So efficient, mechanical means of destroying weeds will eventually be required, along with other practices like intercropping which greatly reduce the need for weed control.
Swarms of these robots will be required. Just to put the efficiency thing in perspective, I can spray herbicide on 120 acres in less than an hour (120' boom at 14 mph). In a day I can spray over a thousand. That's a lot of little robots to do the same job. The nice thing about mechanical (in this case fire!) destruction of weeds is that it's not possible for a plant to be selected for resistance. However there is a possibility for selection based on whatever parameters the AI decides it's looking for. I heard a story not that long about out of China if I recall, where they had a particularly problematic weed that had to be hand picked out of the crop. Eventually the only variants of the weed left looked very much like the crop at that early stage. So I guess even this kind of mechanical control still is problematic.
Anyway it's interesting, particularly how they focused on trying to work with what works for current farms. So many ideas are great, but it's hard to adapt them to the realities I work with, whether its financial, technical, or scale of economy.
Decades ago, before most of you were born, I was learning Tzeltal (a Mayan language of southern Mexico). One day I was out in the corn field with the village headman and his teenage son. The corn plants were a few inches tall, and there were grass blades growing up with them. We were chopping out the grass with machetes. (I know, that seems like using a sledge hammer to... but it's what we had.) I was using my machete to chop something out, when I suddenly realized it was corn instead--the two looked very similar at that stage, I think the grass leaves were slightly bluer. The teenager working next to me saw me almost chop out the corn. "Nopol la'aboj ixim", he said: "You almost cut the corn." It's funny how events like that get cemented in your brain--as I say, it's been decades now, and I've forgotten a lot of the language--but not that.
Anyway, wrt your comment: yes, the grass looked a lot like corn. Whether that was evolution, I don't know; maybe corn (or corn's predecessor, teosinte) and grass started out looking similar millennia ago.
Yes, it won't cause genetic selection as do herbicides (well, maybe it won't, we'll see over time). It will certainly selectively kill some weeds and not others - some have underground stems (e.g. bermuda grass); this will not kill those. For other weeds, I imagine they will need to be careful to adjust the laser to go deep enough to kill the growing point, otherwise the process could select for deeply growing weeds that can recover.
Also, it would be difficult to kill the weeds growing right up against the
Even with proper use, you can't avoid herbicide resistance. Selection pressure is selection pressure. So efficient, mechanical means of destroying weeds will eventually be required, [...]
Selection pressure is selection pressure, regardless of if it’s chemical, thermal, mechanical, or some other force.
The nice thing about mechanical (in this case fire!) destruction of weeds is that it's not possible for a plant to be selected for resistance.
Not true! If the roots go deep enough, the weed may survive, and pretty soon you will have selected for weeds that can survive attacks from lasers. Or perhaps a strain becomes particularly adept at incorporating silica into its structure in such a way that it diffuses the laser, limiting the damage it incurs. There are innumerable variations on this idea.
Rye and oats evolved this way, becoming so much like wheat that they themselves became useful crops! In googling to refresh my memory as to which grain varieties this applied to I learned this is called "Vavilovian mimicry"
https://en.wikipedia.org/wiki/... [wikipedia.org].
Yep, like you say, selection pressure is selection pressure. Kill things that don't look like crops, and the only weeds that survive to pass on their genes will be those that look like crops, with the similarity increasing over time as you get better telling them apart.
As for the relative cost effectiveness of spraying - that could change rapidly if you had to pay for all the health and ecological damage those herbicides do both downstream and to consumers. Right now you're benefiting from exploiting t
Easy to say. Such a criticism of agriculture works both ways, however. You're eating food more cheaply than at any time in history precisely because herbicide use has proven to be relatively safe and effective over a long term, at feeding numbers of people thought impossible a few decades ago. Consequently this monetary cost you want to apply must go all the way back to the price you pay for your food. Are you willing to pay double? Triple? Are you aware of how current farm economics work? I seriously
I think the huge potential value is how this will make low till and no till farming even more productive, by preserving soil health. The robot cannot compete with you using heavy equipment in terms of acreage per hour. But if the robot means you buy zero herbicide, cut your fertilizer costs drastically, and you do not need to buy more expensive genetically modified seed, that is where you win economically. Hopefully, you could be grossing less per acre, and take home more profit.
Definitely the wave of the eventual future (Score:5, Interesting)
Even with proper use, you can't avoid herbicide resistance. Selection pressure is selection pressure. So efficient, mechanical means of destroying weeds will eventually be required, along with other practices like intercropping which greatly reduce the need for weed control.
Swarms of these robots will be required. Just to put the efficiency thing in perspective, I can spray herbicide on 120 acres in less than an hour (120' boom at 14 mph). In a day I can spray over a thousand. That's a lot of little robots to do the same job. The nice thing about mechanical (in this case fire!) destruction of weeds is that it's not possible for a plant to be selected for resistance. However there is a possibility for selection based on whatever parameters the AI decides it's looking for. I heard a story not that long about out of China if I recall, where they had a particularly problematic weed that had to be hand picked out of the crop. Eventually the only variants of the weed left looked very much like the crop at that early stage. So I guess even this kind of mechanical control still is problematic.
Anyway it's interesting, particularly how they focused on trying to work with what works for current farms. So many ideas are great, but it's hard to adapt them to the realities I work with, whether its financial, technical, or scale of economy.
Re:Definitely the wave of the eventual future (Score:5, Interesting)
Decades ago, before most of you were born, I was learning Tzeltal (a Mayan language of southern Mexico). One day I was out in the corn field with the village headman and his teenage son. The corn plants were a few inches tall, and there were grass blades growing up with them. We were chopping out the grass with machetes. (I know, that seems like using a sledge hammer to... but it's what we had.) I was using my machete to chop something out, when I suddenly realized it was corn instead--the two looked very similar at that stage, I think the grass leaves were slightly bluer. The teenager working next to me saw me almost chop out the corn. "Nopol la'aboj ixim", he said: "You almost cut the corn." It's funny how events like that get cemented in your brain--as I say, it's been decades now, and I've forgotten a lot of the language--but not that.
Anyway, wrt your comment: yes, the grass looked a lot like corn. Whether that was evolution, I don't know; maybe corn (or corn's predecessor, teosinte) and grass started out looking similar millennia ago.
Re:Definitely the wave of the eventual future (Score:4, Interesting)
"maize split off from teosinte grass [vox.com] some 9,000 years ago"
Re: (Score:2)
Yes, it won't cause genetic selection as do herbicides (well, maybe it won't, we'll see over time). It will certainly selectively kill some weeds and not others - some have underground stems (e.g. bermuda grass); this will not kill those. For other weeds, I imagine they will need to be careful to adjust the laser to go deep enough to kill the growing point, otherwise the process could select for deeply growing weeds that can recover.
Also, it would be difficult to kill the weeds growing right up against the
Re: (Score:2)
Selection pressure is selection pressure
With this new weed zapper, we're going to breed weeds that look exactly like crops.
Re: (Score:2)
Even with proper use, you can't avoid herbicide resistance. Selection pressure is selection pressure. So efficient, mechanical means of destroying weeds will eventually be required, [...]
Selection pressure is selection pressure, regardless of if it’s chemical, thermal, mechanical, or some other force.
The nice thing about mechanical (in this case fire!) destruction of weeds is that it's not possible for a plant to be selected for resistance.
Not true! If the roots go deep enough, the weed may survive, and pretty soon you will have selected for weeds that can survive attacks from lasers. Or perhaps a strain becomes particularly adept at incorporating silica into its structure in such a way that it diffuses the laser, limiting the damage it incurs. There are innumerable variations on this idea.
Re: (Score:2)
Re:Definitely the wave of the evea hrntual future (Score:2)
The nice thing about mechanical (in this case fire!) destruction of weeds is that it's not possible for a plant to be selected for resistance.
Do you want Doomsday [wikipedia.org]? Because that's how you get Doomsday.
Re: (Score:2)
Yep, like you say, selection pressure is selection pressure. Kill things that don't look like crops, and the only weeds that survive to pass on their genes will be those that look like crops, with the similarity increasing over time as you get better telling them apart.
As for the relative cost effectiveness of spraying - that could change rapidly if you had to pay for all the health and ecological damage those herbicides do both downstream and to consumers. Right now you're benefiting from exploiting t
Re: (Score:2)
Easy to say. Such a criticism of agriculture works both ways, however. You're eating food more cheaply than at any time in history precisely because herbicide use has proven to be relatively safe and effective over a long term, at feeding numbers of people thought impossible a few decades ago. Consequently this monetary cost you want to apply must go all the way back to the price you pay for your food. Are you willing to pay double? Triple? Are you aware of how current farm economics work? I seriously
Re: (Score:2)
I think the huge potential value is how this will make low till and no till farming even more productive, by preserving soil health. The robot cannot compete with you using heavy equipment in terms of acreage per hour. But if the robot means you buy zero herbicide, cut your fertilizer costs drastically, and you do not need to buy more expensive genetically modified seed, that is where you win economically. Hopefully, you could be grossing less per acre, and take home more profit.
Re: (Score:2)
>is that it's not possible for a plant to be selected for resistance.
maybe.
But small parabolic collections of reflecting metal exuded on the surface could be interesting . . . :)
"It's firing back!"
Just as long as the thing recognizes chihuahuas as weeds . . .