No Minister

Posts Tagged ‘Robots

It sure beats a backpack weed sprayer

with 5 comments

Very cool news on the farming weedspraying front.

Farming robot kills 100,000 weeds per hour with lasers

Carbon Robotics has unveiled the third-generation of its Autonomous Weeder, a smart farming robot that identifies weeds and then destroys them with high-power lasers.

As it drives itself down rows of crops, its 12 cameras scan the ground. An onboard computer, powered by AI, identifies weeds, and the robot’s carbon dioxide lasers then zap and kill the plants.

The Autonomous Weeder can eliminate more than 100,000 weeds per hour and weed 15 to 20 acres of crops in one day — for comparison, Myers said a laborer can weed about one acre of his onions per day.

A robot!

With lasers!

Shit, it doesn’t get any better than that and this is the 21st century after all. I’m sure it can be adapted to work the slightly more random world of the average Kiwi sheep, cattle and dairy farm.

Written by Tom Hunter

December 1, 2021 at 12:14 pm

Big Brains.

with 2 comments

The photo shown here is a typical CPU (Central Processing Unit) silicon chip used in desktop and laptop computers.

Specifically this is an AMD Ryzen 3 2200, a so-called “entry-level” chip for people building their own desktop gaming computers. Just a few years ago the power of such chips was still talked of in terms of the number of transistors they held, with one of the classics, the Intel 8086 of the 1980’s, having 29,000.

That AMD chip has 5 billion transistors.

As such people nowadays usually talk about other measurements of power such as clock speed and “cores”. What’s a core? Well it means a processing unit, a computer in itself. If the original silicon chips were said to be “a computer on a chip” then a 4-core chip like the Ryzen 3 2200 has four computers on a chip, and that’s pretty ordinary now. There are retailed chips with 64 cores.

Why do this? Why not just keep making a single core ever larger? Well there are scaling problems, not just in the hardware but in using a single processor to do a job. Instead, use is made of something called parallel computing, where one job is split into many, all run at the same time. It started off as something only used with the supercomputers simulating things like nuclear explosions. Parallel processing makes it possible to perform ever larger data processing jobs in human time.

The thing is that the human brain is also, basically, a parallel processor (Minsky), and a pretty massive one at that, with 100 trillion synapses (brain cells), each of which is like a transistor but with hundreds of connections to other synapses, which multiplies that 100 trillion in terms of data storage and delivers processing power beyond what should be possible given how much slower nerve impulses are compared to electronics.

The SF author Arthur C Clarke, in the book version of 2001: A Space Odyssey, actually makes reference to Minsky’s research on neural networks in explaining how the infamous computer HAL 9000 was developed, which shows you the sort of background study Clarke did for his stories. Minsky was an advisor to the movie.

For decades, most of these neural networks amounted to creating artificial “neurons” in the software, which was basically a clunking simulation, ultimately limited by the hardware it ran on. You could do interesting things, just slowly.

Which brings me to this news story, World’s Largest Chip Unlocks Brain-Sized AI Models With Cerebras CS-2.

Cores? It has 850,000 of them!

Cerebras Systems today announced that it has created what it bills as the first brain-scale AI solution – a single system that can support 120-trillion parameter AI models, beating out the 100 trillion synapses present in the human brain. In contrast, clusters of GPUs, the most commonly-used device for AI workloads, typically top out at 1 trillion parameters. Cerebras can accomplish this industry-first with a single 850,000-core system, but it can also spread workloads over up to 192 CS-2 systems with 162 million AI-optimized cores to unlock even more performance. 

Are you scared?

I am, and I’ve been in this world for most of my life.

Of course it’s going to take some time to develop the software to truly use this baby, but if we get to the point where the software-hardware configuration can start truly learning in growing itself, well…

Written by Tom Hunter

October 7, 2021 at 9:07 am

Funny … and scary?

leave a comment »

Things are perhaps getting too depressing and heated.

So it’s time for a KitKat moment.

I’m putting both of these videos up together with a simple question for readers.

Rating both videos, are they:

a) Scary

b) Funny

c) Funny and Scary

d) Neither Funny nor Scary

First up is “Inside Biden’s Office”:

Next up is our (slightly?) more distant future. You may have seen YouTube clips published by a robotics company called Boston Dynamics, showing off the robots they’re developing for various uses, starting with “pack dogs” for the US military. They are eerily like animals, and their humanoid robots increasingly move like humans.

Those videos have led to this parody – at least I hope it’s a parody. I’ve never had a lot of faith in the idea of built-in, programmed restrictions on robotic brains, starting with Asimovs Three Laws of Robotics.

Still, in this case I can’t blame the robot at all.

That’s what scary about it. 🙂

Written by Tom Hunter

August 21, 2021 at 1:26 pm

Well, this is depressing

with 4 comments

No need for this level of complexity

Specifically the news that the race is on to build killer robot armies.

They won’t look anything like James Cameron’s famous images from his dystopian hell of The Terminator movies.

(By the way, watch only the first two of the series. After the 1991 sequel they’re totally derivative crap designed only to pull money from your wallet, a warning from friends that I had already guessed at as I avoided them.)

Blonde and here to kill you.

Still less is it going to look like the Cylons in Battlestar Gallactica (BSG) such as “Six”, more’s the pity.

No, as is often the way of reality vs fantasy they’ll look a lot more mundane, probably not too different to the sort of drones you can buy off-the shelf nowadays.

And that’s what really frightening about them. Unlike nuclear weapons it doesn’t take a lot of infrastructure or resources to build large numbers of these things.

Also, don’t imagine that an “AI killer robot” is going to have some sort of human-level intelligence, or need to.

That’s not what Artificial Intelligence is really about, despite decades of SF stories like BSG.

The “AI” in this case will amount to little more than the ability to do the following:

  • Recognise a human target, which could be just any human or perhaps using facial or body recognition (or your cellphone)
  • Control flight and/or other movements towards the target.
  • Trigger a lethal munition to kill the target. Lethal meaning something as small as a single bullet.

It should be noted that all these capabilities are here now.

The temptation to open Pandora’s Box is irresistible. In early March, the U.S. National Security Commission (NSC) on Artificial Intelligence completed its two-year inquiry, publishing its findings in a dense 750-page report. Its members unanimously concluded that the United States has a “moral imperative” to pursue the use of lethal autonomous weapons, a.k.a. “killer robots.” Otherwise, we risk bringing a rusty knife to a superhuman gunfight.

Citing the threat of China or Russia leading the global artificial intelligence (AI) arms race, the commission’s chairman, former Google CEO Eric Schmidt, urged President Biden to reject a proposed international ban on AI-controlled weapons. Schmidt rightly suspects our major rivals won’t abide by such a treaty, warning U.S. leaders, “This is the tough reality we must face.”

If other superpowers are going to unleash demonic drone swarms on the world, the logic goes, the United States should be the first to open the gates of Hell.

Of course we already have things like the General Atomic Predator drones (“General Atomic”, how 1950’s is that?) and others which have been launching missiles at people for over a decade now. But they have humans in the decision loop and they’re still big and relatively expensive, although much cheaper than a human-piloted fighter bomber.

The attack drones currently on the market are plenty dangerous as is. A good example is the KARGU Loitering Munitions System, currently deployed by Turkish forces. This lightweight quadcopter “can be effectively used against static or moving targets through its … real-time image processing capabilities and machine learning algorithms.”

KARGU’s mode of attack is full-on kamikaze. It hovers high in the air as the operator searches for victims. When one is located, the drone dive-bombs its target and explodes. If the concussion doesn’t kill them, the shrapnel will. Just imagine what a thousand could do.

That last is the future. What we’re talking about here is a swarm of such machines and again – not like SF – these don’t need any centrally organised intelligence, human or AI, to operate. For twenty years now computer simulations have mimicked the swarming movements of schools of fish and flocks of birds with just three rules.

Once you get into such swarms we’re no longer talking about just picking off a few selected targets:


To raise awareness of this imminent threat, the Future of Life Institute produced the alarming, if poorly acted film Slaughterbots. The finale shows dissident college students having their brains blown out by bird-sized quadcopters.

In a 2018 study conducted for the US Air Force, drone specialist Zachary Kallenborn correctly argued that lethal drone swarms should be declared weapons of mass destruction.

Cheap weapons of mass destruction, too.

Even without that miserable conclusion from the USNSC I would have found it hard to believe that various nations could be held back from pursuing development of these things.

In the future how tempted would a POTUS be by the idea that the entire North Korean nuclear team, military and scientists, could be taken out in one hit by such a swarm, leaving nobody to launch a nuclear counter-strike? Or imagine an Israeli leader looking at the Iranian nuclear group? And that’s in democratic nations. What brakes might there be on the likes of Xi Jinping, Putin and Erdogan?

Of course every weapon system has been countered sooner or later. In this case it may be that in future we’ll each be guarded by a small swarm of counter-drones, starting with the wealthy members of society like Eric Schmidt:

In 2019, PAX published a list of the global corporations most likely to develop lethal autonomous weapon systems. Among the U.S. companies ranked as “high risk” are Amazon, Microsoft, and Oracle, as well as Intel, Palantir, Neurala, Corenova, and Heron Systems. It’s worth noting that the top members of the National Security Commission on AI—all of whom support using these murder machines—include chiefs from Amazon, Microsoft, and Oracle.

Written by Tom Hunter

July 15, 2021 at 8:13 am

The 21st Century Space Race

with 2 comments

The news on April 16 that NASA has selected the SpaceX Starship design for two missions to the Moon was stunning enough on its own, but the last few months have seen some other important events that show that space exploration, including human exploration, is starting to rapidly increase its tempo.

April was an especially busy month.

There have been two spectacular robotic missions to Mars. The third SpaceX crew mission to the ISS was launched. Another four astronauts returned in their original SpaceX Dragon craft. The Chinese have launched the first module of their new space station into orbit, sent Taikonauts to it and landed their first Mars probe.

MARS

Yes, that is a drone helicopter on Mars.

What you are looking at in these pictures is the first flying machine on another planet. This is the spacecraft, Ingenuity, a little helicopter weighing only a couple of kilograms, and these pictures are from its 4th test flight, conducted on April 29. The first was April 19.

It was only expected to fly a few missions as a pure test vehicle, starting with a simple up-and-down flight, followed by moving around the landscape, returning to where it started and so forth.

However, like many other American robots of recent years, it has proved far more durable than expected, and since it provides the ability to scout ahead of the rover Perseverance (its “mother ship” that landed on February 18) the JPL team has decided to use it for that until it eventually fails.

Planning the route that such rovers have to drive is a slow and careful process; some years ago one of the solar-powered rovers, Opportunity, got stuck in an unseen sand trap. Ingenuity is a huge help in preventing that.

That map plan was from the 9th planned flight, which was completed a couple of days ago, with the machine flying a distance of 2,051 feet (625 meters) at 5 meters (16 feet) per second and remaining airborne for approximately 2 minutes, 47 seconds. They’re really pushing the little beast beyond its limits:

The onboard algorithm which lets Ingenuity determine where it is along the flight path, was designed for a comparatively simple technology demonstration over flat terrain and does not have the design features to accommodate high slopes and undulations that are to be found in Séítah. 

You can keep track of the flights at Robert Zimmerman’s Behind The Black blog.

However, the USA has finally been matched in the Mars Rover department when, on May 14, China’s Tianwen-1 spacecraft landed and on May 22 sent its Zhurong rover trundling onto the surface. With its solar panels it looks awfully similar to the Opportunity and Spirit rovers (see An Everlasting Itch For Things Remote) sent to Mars by NASA in the 2000’s. It will be interesting to see if it lasts as long as they did.

SpaceX and the ISS

On April 23rd SpaceX launched a crew of four to the ISS in the company’s second such mission, called Crew-2. The key mission first was that both the Falcon 9 rocket and the Dragon spacecraft were reused machines. The Crew Dragon, named Endeavour, had flown the historic Demo-2 mission (their first crewed flight) and the Falcon 9 rocket had pushed the Crew-1 astronauts to the ISS last year, in the first operational flight.

They crossed paths with the Crew-1 team which also consisted of four astronauts, making things rather crowded on the ISS for a few days before they returned to Earth on May 2nd. As you can see from the link, these splashdowns are rather more casual than the days of Apollo when countless US Navy ships were deployed far out in the Pacific. Here, they’re only just offshore from the Florida coast, and SpaceX kept the exact landing spot quiet this time so that they would not be surrounded by rubber-necking boaties as happened for the return of Demo-2.

Space X is contracted for four more crewed missions but will likely get more than that as the ISS is expected to be up there for a few more years: 2024 has been discussed as a shutdown date, with the Russians perhaps quitting then. However, new modules continue to be launched, the station is working well and since there’s no replacement it seems silly to abandon it even if its equipment is starting to get aged and obsolete.

As a side note, one continuing disappointment has been Boeing’s CST 100 Starliner spacecraft – pictured on the right – which was supposed to have been on the same schedule as SpaceX’s Dragon so that NASA could have a backup. Instead it’s fallen badly behind schedule but will attempt another uncrewed test flight at the end of July.

China’s space station and Russia

Another reason the USA and ESA may hang on to the ISS for longer than planned, is that on April 29, just days after its Mars probe landing, China launched the first module of its new Taingong space station, followed by three “Taikonauts” on June 20.

Since then they’ve been busy conducting spacewalks, readying everything for the next modules to be sent up. The plan is to have it finished by the end of 2022. It will be about 1/5 the size of the ISS, comparable to the old Russian Mir space station.

By the time Mir crashed into the Pacific Ocean in 2001, Russia had already been hard at work for a decade with NASA on the ISS, just one part of the whole post Cold War effort by the USA to shovel money into Russian technology centres to keep their scientists and engineers in-country and not working with the likes of North Korea.

But the world has changed. Between losing the lucrative ISS passenger business to SpaceX and with their announced plan to stop using the ISS by 2024, Russia was clearly removing itself somewhat. Still, the US is proceeding into its post-ISS future with the Artemis programme to land astronauts on the Moon and establish a base there by the end of this decade and Russia could well have been expected to join that. So it came as a bit of shock in March when Russia announced an agreement with China to build a lunar base together, the International Lunar Research Station (ILRS). No schedule has been announced but Russia clearly feels more comfortable in China’s orbit than in the US’s.

The US Military and SpaceX

A few weeks ago the US Air Force released a 462-page report detailing how it intends to spend its $200 billion budget. Such is routine bureaucracy, but on page 305, under the heading of “Rocket Cargo” was a very interesting section:

“The Department of the Air Force seeks to leverage the current multi-billion dollar commercial investment to develop the largest rockets ever, and with full reusability to develop and test the capability to leverage a commercial rocket to deliver AF cargo anywhere on the Earth in less than one hour, with a 100-ton capacity,

Although no company or vehicle was mentioned by name, there is only one vehicle even on the drawing boards that has that capability.

It will likely be a decade before this comes to fruition, as the USAF will want to see a lot of successful Starship cargo flights, including sub-orbital hops around the Earth, before it will start laying contracts to buy flights and possibly even order Starship variants to its specifications. In this it would be following in the wake of the Boeing 747, which was built with military as well as commercial use in mind.

Written by Tom Hunter

July 12, 2021 at 6:00 am