Tag Archives: analysis

Grapes of Metallic Wrath?

Posted on by

There are words, such as ‘freedom’ and ‘democracy’ which get tossed around like a cat’s toy, tumbling between meanings depending on the speaker. To this list of floating signifiers, we can add ‘automation’ which, when mentioned by business types, is meant to depict a bright and shining future but, when used by people on the left who, one would hope, are concerned about the prospects for labor, is typically employed as a warning of trouble ahead (there’s an exception to this: the ‘fully automated luxury communism’ folks, some of whom, seeing their robot butler dreams fade, are now turning to the polar opposite idea of degrowth).

The trouble with floating signifiers is that they float, making it difficult to speak, and perhaps think, with precision – actually, this also explains their appeal; any malcontent can shout they’re defending ‘freedom’ and fool at least some of the people, some of the time, via a socially agreed-upon vagary.

One of my quixotic preoccupations is a struggle against imprecise language and thought. It’s silly; we’re all over the place as a species and wouldn’t be human if it were otherwise (among my many arguments against the ‘AI’ industry crowd is its collective failure to understand that imprecision is a key element of our cognition, beyond duplication in electronic machinery)

So, with my quest for precision in mind, let’s spend a few moments contemplating automation, trying to put some bones and flesh on an ideological mist.

Check out the graphic shown below:

The Things to Think About and Study

I cooked up this image to visualize what I see as the appropriate areas of material concern for left politics. How do things work? And, for me, because this is my area of expertise, what role does computation play in the performance and command and control of labor in these various sectors of production?

In this post, I focus on automation in farming. Oh and by the way, my focus here is also on method, on how to think; that is, how to think in material terms about things which are presented in vague ways. 

Drones, Robot Tractors and Harvestors 

For me, the foundational, 21st century work on the real-world impacts of automation on labor is ‘Automation and the Future of Work’ by Aaron Benanav. Here’s a link to an article Benanav wrote for the New Left Review outlining his argument which can be summarized as: yes, of course, there’s automation and it has an impact but not as profound and far reaching, and not in the ways we are encouraged to think.

To look at farming specifically, I visited PlugandPlay, an industry and venture capitalist boosting website (trade publications, properly analyzed, are an excellent source of information) that published “How Automation is Transforming the Farming Industry”. 

From the article:

“Farm automation, often associated with “smart farming”, is technology that makes farms more efficient and automates the crop or livestock production cycle. An increasing number of companies are working on robotics innovation to develop drones, autonomous tractors, robotic harvesters, automatic watering, and seeding robots. Although these technologies are fairly new, the industry has seen an increasing number of traditional agriculture companies adopt farm automation into their processes.”

https://www.plugandplaytechcenter.com/resources/new-agriculture-technology-modern-farming/

You can imagine a futuristic farm, abuzz with robotic activity, all watched over, to paraphrase the poet Richard Brautigan, by machines of sublime grace, producing the food we need while the once over-worked farmer relaxes in front of a panel of screens watching devices do all the labor.

Let’s dig a little deeper to list the categories of systems mentioned in the article:

  • Drones
  • Autonomous tractors
  • Robotic harvesters
  • Automatic watering
  • Seeding robots

For each of these categories, the PlugandPlay article, very helpfully, provides an example company. This gives us an opportunity to review the claims, methods and production readiness (i.e., can you buy a product and receive shipment and technical support for setup or are only pre-orders available?) of individual firms in each area of activity. This information enables us to add more precision to our understanding.

With this information at-hand, we’re not just saying ‘farming automation’ we’re looking at the sector’s operational mechanics.

For drones, American Robotics’ aerial survey systems are mentioned. As is my habit, I checked out their job listings to see the sort of research and engineering efforts they’re hiring for which is a solid indicator of real or aspirational capabilities. I’ve written about drone-based analysis before; it does have real world applications but isn’t as autonomous as often claimed.

The three examples of robotic harvesters listed are from Abundant Robotics, which is building specialized apple harvesting systems, Bear Flag Robotics, which seems to have retrofitted existing tractors with sensors to enable navigation through farming fields (and perhaps remote operation, the marketing material isn’t very clear about this) and Rabbit Tractors, which appears to be out of business.

There are a few other examples offered but hopefully, a picture is forming; there are, at this point, some purpose built systems – some more demonstration platform than production ready – which show the limitations, and potential usefulness of automation in the farming sector: perfect for bounded, repetitive applications (a weed sprayer that follows assigned paths comes to mind) not so great at situations requiring flexible action. Keep this principle in mind as a rule of thumb when evaluating automation claims.

It also isn’t clear how well any of these systems work in varying weather conditions, what the failure modes and maintenance schedules are and lots of other critical questions. It may seem cheaper, in concept, to replace workers with automated or semi-automated harvesters (for example) but these machines aren’t cheap and introduce new cost factors which may complicate profitability goals and it follows, adoption by agribusiness, which, like all other capitalist sectors, is always in search of profits.

…

So, yes, automation is indeed coming to, or is already present in farming but not, it appears, in the hands-off, labor smashing way we tend to think of when the word, ‘automation’ is tossed around, like a cat’s toy.

Next, time, I’ll take a look at automation in logistics. How far has it gone? How far will it go? 

AI Ethics, a More Ruthless Consideration

Posted on by

According to self-satisfied legend, medieval European scholars, perhaps short of things to do compared to we ever-occupied moderns, spent countless hours wondering about topics such as, how many angels could simultaneously occupy the head of a pin; the idea being that, if nothing was impossible for God, surely, violating the observable rules of space and temporality should be cosmic child’s play for the deity…but to what extent?

How many angels, oh lord?

Although it’s debatable whether this question actually kept any monks up at night more than, say, wondering where the best beer was, the core idea, that it’s possible to get lost in a maze of interesting, but ultimately pointless inquiries (a category which, in an ancient Buddhist text is labeled, ‘questions that tend not towards edification’) remains eternally relevant.

At this stage in our history, as we stare, dumbfounded, into the barrels of several weapons of capitalism’s making – climate change being the most devastating – the AI endeavor is the computational equivalent of that apocryphal medieval debating topic; we are discussing the ethics of large language models, focusing, understandably, on biased language and power consumption but missing a more pointed ethical question: should these systems exist at all? A more, shall we say, robust ethics would demand that in the face of our complex of global emergencies, tolerance for the use of computational power for games with language cannot be justified.

OPT-175B – A Lesson: Hardware

The company now known as Meta recently announced its creation of a large language model system called OPT-175B. Helpfully, and unlike the not particularly open OpenAI, the announcement was accompanied by the publication of a detailed technical review, which you can read here.

As the paper’s authors promise in the abstract, the document is quite rich in details which, to those unfamiliar with the industry’s terminology and jargon, will likely be off-putting. That’s okay because I read it for you and can distill the results to four main items:

  1. The system consumes almost a thousand NVIDIA game processing units (992 to be exact, not counting the units that had to be replaced because of failure)
  2. These processing units are quite powerful, which enabled the OPT-175B team to use relatively fewer computational resources than what was installed for GPT-3 another, famous (at least in AI circles) language model system
  3. OPT-175B, which drew its text data from online sources, such as that hive of villainy, Reddit, has a tendency to output racist and misogynist insults
  4. Sure, it uses fewer processors but its carbon footprint is still excessive (again, not counting replacements and supply chain)

Here’s an excerpt from the paper:

“From this implementation, and from using the latest generation of NVIDIA hardware, we are able to develop OPT-175B using only 1/7th the carbon footprint of GPT-3. 

While this is a significant achievement, the energy cost of creating such a model is still nontrivial, and repeated efforts to replicate a model of this size will only amplify the growing compute footprint of these LLMs.” [highlighting emphasis mine]

https://arxiv.org/pdf/2205.01068.pdf

I cooked up a visual to place this in a fuller context:

Here’s a bit more from the paper about hardware:

“We faced a significant number of hardware failures in our compute cluster while training OPT-175B. 

In total, hardware failures contributed to at least 35 manual restarts and the cycling of over 100 hosts over the course of 2 months. 

During manual restarts, the training run was paused, and a series of diagnostics tests were conducted to detect problematic nodes.

Flagged nodes were then cordoned off and training was resumed from the last saved checkpoint. 
Given the difference between the number of hosts cycled out and the number of manual restarts, we estimate 70+ automatic restarts due to hardware failures.”

https://arxiv.org/pdf/2205.01068.pdf

All of which means that, while processing data, there were times, quite a few times, when parts of the system failed, requiring a pause till fixed or routed around (resumed, once the failing elements were replaced).

Let’s pause here to reflect on where we are in the story; a system, whose purpose is to produce plausible strings of text (and, stripped of the obscurants of mathematics, large-scale systems engineering and marketing hype, this is what large language models do) was assembled using a small mountain of computer processors, prone, to a non-trivial extent, to failure.

As pin carrying capacity counting goes, this is rather expensive.

OPT-175B – A Lesson: Bias

Like other LLMs, OPT-175B has a tendency to return hate speech as output. Another excerpt:

“Overall, we see that OPT-175B has a higher toxicity rate than either PaLM or Davinci. We also observe that all 3 models have increased likelihood of generating toxic continuations as the toxicity of the prompt increases, which is consistent with the observations of Chowdhery et al. (2022). As with our experiments in hate speech detection, we suspect the inclusion of unmoderated social media texts in the pre-training corpus raises model familiarity with, and therefore propensity to generate and detect, toxic text.” [bold emphasis mine]

https://arxiv.org/pdf/2205.01068.pdf

Unsurprisingly, there’s been a lot of commentary on Twitter (and no doubt, elsewhere) about this toxicity. Indeed, almost the entire focus of ‘ethical’ efforts has been on somehow engineering this tendency away – or perhaps avoiding it altogether via the use of less volatile datasets (and good luck with that as long as Internet data is in the mix!)

This defines ethics as being the task of improving a system’s outputs – a technical activity – and not a consideration of a system as a whole from an ethical standpoint within political economy. Or to put it another way, the ethical task is narrowed to making sure that if I use a service which, on its backend, depends on a language model for its apparent text capability, it won’t in the midst of telling me about good nearby restaurants, hurl insults like a klan member.

OPT-175B – A Lesson: Carbon

Within the paper itself, there is the foundation of an argument against this entire field, as currently pursued:

“...there exists significant compute and carbon cost to reproduce models of this size. While OPT-175B was developed with an estimated carbon emissions footprint (CO2eq) of 75 tons,10 GPT-3 was estimated to use 500 tons, while Gopher required 380 tons. These estimates are not universally reported, and the accounting methodologies for these calculations are also not standardized. In addition, model training is only one component of the over- all carbon footprint of AI systems; we must also consider experimentation and eventual downstream inference cost, all of which contribute to the growing energy footprint of creating large-scale models.”

A More Urgent Form of Ethics

In the fictional history of the far-future world depicted in the novel ‘Dune’ there was an event, the Butlerian Jihad, which decisively swept thinking machines from galactic civilization. This purge was inspired by the interpretation of devices that mimicked thought or possessed the capacity to think as an abomination against nature.

Today, we do not face the challenge of thinking machines and probably never will. What we do face however, is an urgent need to, at long last, take climate change seriously. How should this reorientation towards soberness alter our understanding of the role of computation?

I think that, in face of an ever-shortening amount of time to address climate change in an organized fashion, the continuation, to say nothing of expansion of this industrial level consumption of resources, computing power, talent and the corresponding carbon footprint is ethically and morally unacceptable.

At this late hour, the ethical position isn’t to call for, or work towards better use of these massive systems; it’s to demand they be halted and the computational capacity re-purposed for more pressing issues.  We can no longer afford to wonder how many angels we can get to dance on pins.