robert smith

There’s no knowledge in chat GPT…

Chat GPT is an epistemic agent

Actually good AI images

The first good analysis of language models I’ve seen.

more 99%I teaching content on bicycles.

99% invisible does it again – a story about people representing wild rice in the US legal system.

Interview with Vauhini Vauna who is a tech writer. She wrote an essay collaboratively with the open AIs GPT3. End of the interview touches on this, and particularly the affective dimensions of working with AI. It would be a good unexpected teaching case.

Sensitive podcast on American’s relationship with an endangered species. Takes a kind of socio-technical systems / actor-centred approach to Grouse as an issue.

The default metaphor for intellectual property in modern times is “ownership.” In this model of ownership, all ideas, stories, inventions, characters, product names, techniques are understood to be inherently born as the property of their creator. These thoughts-made-real are seen to be owned by the mind that births them. You think them, you own them. With this status of ownership, intangible creations such as a novel, a musical melody, a plot, a phrase, formula, etc — all things created by a mind — are given a monopoly of rights in order to encourage further creations by the same creator. And to spur others to create. This lawful monopoly — such as copyright, patents, trademarks — protects the creation from being used by others for gain. By current law, this inherent monopolistic ownership is held strongly for long periods of time, ranging from decades to a century, depending on the conceptual type (patents may be 17 years while copyright may exceed lifetimes). This awarded monopoly has a few exceptions for very limited special cases, such as “fair use” and public domain. In these modes anyone can fairly use the invention for their own purposes. Certain restrictions may apply, like if the use might need to be for education, or for parody, or so used in a transformative way, or bettered by the use. These exceptions were to be kept to an absolute minimum in order to maximize the monopoly of the hard working creator. This framing plays into both the modern idea of ownership as the sacred foundation of wealth and prosperity, but it also plays into the idea of creator as a hero, or at least as the bedrock of progress.

Big tech companies’ research budgets are massive

An inventor finally offers an interesting take on responsibility, centring on narratives and the ability to talk about the ambivalences of NFTs.

Podcast with Max Chafkin, author of a book on Peter Thiel. Has a really nice bit of emerging tech disassembly on a scary technology (Palantir) and an example of how the wrong assessment of some tech visionaries about a problem (9/11 – failure of surveillance) led to its development.

Surreptitious data sharing mental health apps

Uranium mining in the USA

Economist article on Google releasing petition folding predictions

Government creates a new science and technology council, chaired by the prime minister, to make decisions about research to address societal challenges.

Great case study of funding and technological change and its imbrication in politics (it made Britain better at being Imperialistic)

Profile of a bioengineering person who doesn’t think in terms of genes and is trying to drive regeneration using bioelectrical signals. Very reminiscent of Karen barads talks

Academic title = The role of digital technologies in changing international politics: the case of diplomacy

How collaboration shapes knowledge production: "A 2009 study showed that about 80 per cent of Central Africa’s research papers were produced with collaborators based outside the region. With the exception of Rwanda, each of the African countries principally collaborated with its former coloniser. As a result, these dominant collaborators shaped scientific work in the region. They prioritised research on immediate local health-related issues, particularly infectious and tropical diseases, rather than encouraging local scientists to also pursue the fuller range of topics pursued in the West."

Amazing article about how AirPods are evil

Not synthetic biology

Somatosphere forum on the origins of the coronavirus responsible for this pandemic.

You could imagine a future in which the pandemic drives vaccine development which negates the need for eradicating mosquitoes from the world.

More future organisms from the BBC

Dazed and confused is a cool magazine. This is from 2013 and is basically all arty tropes.

James Somers on how an English professor in the US (10 years ago) dealt with student feedback – by emailing with students as they go through the course.

Steve Jobs
February 6, 2007

Apple has a long relationship with Adobe. In fact, we met Adobe’s founders when they were in their proverbial garage.

Journalists doing stuff that academics haven’t with data and tracking, again.

Aside from the fact that this is a really nice policy report between science and the humanities, it contains this quote for anyone following new / emerging / novel things: “Professor Agar explains how data-driven decisions are older than commonly thought, and how concerns about privacy and automation go back much longer than any data revolution. He emphasises that policymakers focus too much on governing new technologies, when it is often the older ones that really matter.”

Anti Ice-9!

Contra this, ‘the people tend to be one or two rings down, who have actually spent the time creating technologies and now find themselves trying to extract themselves from it’. > “It may or may not be relevant that Rosenstein, Pearlman and most of the tech insiders questioning today’s attention economy are in their 30s, members of the last generation that can remember a world in which telephones were plugged into walls. It is revealing that many of these younger technologists are weaning themselves off their own products, sending their children to elite Silicon Valley schools where iPhones, iPads and even laptops are banned. They appear to be abiding by a Biggie Smalls lyric from their own youth about the perils of dealing crack cocaine: never get high on your own supply.”

A couple of things interesting to STS: 1. The question about ethics and political power of the technology he builds becomes one of whether people are comfortable about acceleration. 2. ‘Democratise’ here means spread as far as possible. Of course they’re just snippets of an interview transcript but they seem familiar patterns. And of course, he is the CEO. It would be unprecedented if he acknowledged what a lot of people see as a critical position had any validity.

“This can be achieved by, for example, using open processes in the planning stage of power plants that allow the public to comment on and influence the terms and consequences of models, or by including forms of international peer review for national modeling assumptions.” This is one of my discomforts with my own work and others in the field. That we are essentially saying, when we make recommendations like this, that such processes are less likely to a) fail or b) produce catastrophic failure if they do. But this is even less testable than the accuracy of models and comes with a significant burden. The next sentence does suggest a way forward — finding ways to protect the poor — but this is not a technologically-specific solution but a question of equity.

"Now it is the turn of artificial intelligence, nanotechnology, biotechnology, materials science, 3D printing and quantum computing to transform the global economy."

A suggestion from a guardian columnist about how to deal with megacompanies. Note the reference to knowing the future in the Nature paper. Interesting.