Newsletter No.282 — Oct 08, 2023

Against Scale ⊗ AI Data Training Companies Like Scale AI Are Hiring Poets ⊗ Daniel Schmachtenberger on the Metacrisis

Join Sentiers, the futures thinking observatory

Also this week → How civic infrastructure creates shared prosperity ⊗ Eudaimonia machine and professional time ⊗ The museum of the future of productive forces ⊗ Building an observatory

Against scale

Last week I shared an interview with Mustafa Suleyman where he referred to synthetic biology and AI as “life and intelligence.” It felt more like two things evolving in parallel, two main topics that will influence society going forward. One could see all of Claire L. Evans’ writings at Grow as diving much deeper into those two topics and intersecting human creations, like computation and industrialisation, with nature’s various intelligences and growth.

Here Evans argues against the concept of scalability, which she posits as a social and economic construction that is alien to both biological and material reality. Evans argues that scalability ignores the negotiations between individual elements and is predicated on hierarchy, isolation, and the commodification of nature. Instead, synthetic biology should model itself on the living world, which takes a collaborative approach to survival and allows for fruitful cross-contamination of ideas.

Evans mentions that scalability assumes that “nature is little more than a raw material to be processed and commodified until it is spent.” It made me think that perhaps we should instead/also talk of ‘raw industry’; the unthinking use of nature.

In the name of progress, Tsing observes, we call such expansion “growth,” as though we were speaking of something alive. But it’s not alive; it’s a cousin of death, and it has made life, with its nonscalable elements — poppies, mushrooms, stubborn bacteria — seem like an impediment, a spanner in the works grinding the great churn of expansion to a halt. […]

It’s only by shunting away responsibility for externalities like waste, the physical and mental health of workers, or the depletion of nonrenewable resources, and by stripping context from all its component parts that an enterprise can be made scalable to begin with. […]

The living world simply does things well, and it’s often easier to crib from the useful properties of natural systems than it is to start from scratch. […]

In synthetic biology, the way forward may not be a matter of producing at scale but rather inquiring at scale, changing the volume at which we converse with the living world before deciding what to assemble, rather than mine, maul, or murder. […]

Although it can often feel as though we live on the knife’s edge between the inconceivably large — climate change, big data — and the vanishingly small — deadly viruses, toxic particles — scale is not linear. I am a tumult of cells and bacteria; I am a speck of dust in the cosmos; I am 5’9” and walking along the trail towards a glowing orange poppy field, all at once.

AI data training companies like Scale AI are hiring poets

Poets, novelists, and writers with advanced degrees are being hired to provide companies training LLMs with high quality training data. Much like actors getting their body, gestures, and voice synthesised for future use, I’m sure these are one-shot deals, that results will be used in perpetuity, and that the authors will get nothing but salaries during the time they provide the writing. People in the global south have already been tagging training data for a while and have been getting paid peanuts, so this isn’t new, just a different form. Admittedly, all of us are actually doing the same thing for free by using all of these AI products. Are you using ChatGPT or is it using you?

“They are trained to reproduce. They are not designed to be great, they try to be as close as possible to what exists,” Fabricio Goes, who teaches informatics at the University of Leicester, told Rest of World, explaining a popular stance among AI researchers. “So, by design, many people argue that those systems are not creative.” […]

[Various companies and artists] claim their copyrighted work was included in ChatGPT’s training data without permission, since the tool can accurately summarize their work and imitate their style. Any text written for Scale AI or Appen, however, is likely to be owned in full by the training data company or its clients.

More → Benedict Evans looks at AI and the automation of work and argues that throughout history, automation has led to the creation of new jobs and is likely to do so once again. The development and arrival of AI is happening much quicker than other waves of automation, and while it may take time to fully integrate these technologies into the workplace, Evans believes that they have the potential to create new opportunities and efficiencies.

The tools that people use to do their jobs, and the tasks that might now get a new layer of automation, are complicated and very specialised, and embody a lot of work and institutional knowledge. A lot of people are experimenting with ChatGPT, and seeing what it will do. If you’re reading this, you probably have too. That doesn’t mean that ChatGPT has replaced their existing workflows yet, and replacing or automating any of those tools and tasks is not trivial. […]

While one could suggest that LLMs will subsume many apps on one axis, I think it’s equally likely that they will enable a whole new wave of unbundling on other axes, as startups peel off dozens more use cases from Word, Salesforce and SAP, and build a whole bunch more big companies by solving problems that no-one had realised were problems until LLMs let you solve them.

§ If you want to develop a much better understanding of the world, I suggest these two things I’ve been working through over the week. Caveat; you might find the conclusions depressing, but that’s addressed in the first. To start, galaxy-brained Daniel Schmachtenberger spoke about the Metacrisis at the Stockholm Impact/Week (Video, via Rabbit Holes). Then, if you read French, run to a copy of Jean-Marc Jancovici and Christophe Blain’s Le Monde Sans Fin. If you don’t, be sure to pre-order the translation, World Without End. Brain explosions galore.

§ How civic infrastructure creates shared prosperity. Basically, instead of giving incentives for businesses to move in, just make cities better places to live, the rest will follow. “Less than 2 percent of [entrepreneurs] surveyed mentioned business incentives, but many mentioned connections to people and quality spaces. ‘Companies that grow are led by people who could live anywhere,’ said Morris. ‘People who can be anywhere want to live in great places.’” […] “We often focus on the cost, but not the value of public space investments, what makes a great place is the value of that place to people, but too often, those who make the decisions focus only on the cost.”

§ As I mentioned on LinkedIn, a couple of people over the last week or two have mentioned an idea they discovered on the newsletter, the Eudaimonia machine, so I thought I’d re-share the article. Fluid work, office spaces with different areas for different concentration and collaboration needs, spaces shared by multiple companies at different moments, flipped workplaces (referring flipped classrooms), etc. Related, I quite enjoyed Sahil Bloom’s four types of professional time; management, creation, consumption, and ideation (this last one is likely not what you’re thinking). Not sure how one might actually put the four in place as an employee or partner but a useful framework.

§ I’ve been dabbling with the idea of ‘topic explorations’ as a framework for Sentiers, especially memberships. A kind of metaphorical mix of Craig Mod’s walks, Les Grands Explorateurs, or National Geographic; going on expeditions and reporting back. No surprise then that I really love Meg Conley’s vision for an observatory. “The people in these observatories make records of what they see, hear and detect. One observation on its own is a curiosity. But enough observations, from enough people, over enough years can help map a corner of existence. I am a fan of these observatories! And we need more of them.”

Futures, foresights,
forecasts & fabulations

The museum of the future of productive forces
“The Museum of the Future is futuristic in this way. It’s less about a future in which new secrets of the universe have been unlocked by science, and more about one in which today’s knowledge and tech are deployed by more advanced productive forces to accomplish more ambitious—and more abstract—goals.”

Sailing the seas of superdensity
This week Scott Smith mentioned it’s been ten years since the publication of the above piece. Still a must read.

Aperture: An approach for transformative futures & foresight
This one by john a. sweeney is quite a bit denser and academic than what I usually share here. If you are digging into foresight and futures there are a lot of references in there to different views and methods, you can use the piece as a map of the field.