timelets | Entries tagged with process

Time to think beyond Trump, but to do that we need a key data point — the process, not the outcome, of the 2026 elections. If Trump manages to intimidate voters, politicians, business people and judges to subvert the elections, we are going to have a completely different country than the democracy we've had so far. The next 10 months are going to define the next 10+ years and beyond. So far, only the markets passed the stress test.

It’s hard to know what Mr. Trump might do next, which feeds public anxiety. But as his popularity ebbs, so does his political capital. His approval rating has sunk, his mass deportations are seen as excessive, tariffs are unpopular, and even GOP voters disliked his Greenland demands. Democrats took November’s races in Virginia and New Jersey in a rout. The GOP House majority is in peril, and the Senate is competitive. Mr. Trump’s attempts to gerrymander a safer House majority have backfired as Democrats have done the same.

The ultimate check on power is an election, and on that score Mr. Trump’s bull-dozing governance may be building the opposition that costs his party its majority in November.

https://www.wsj.com/opinion/donald-trump-greenland-congress-supreme-court-restraints-84dbc858

There's a growing understanding in the field that producing human-like texts does not imply human-like cognitive processes. Using the traditional terms like "Artificial Intelligence", "Neural Networks", etc. obscures that fact. (I wish I could up with a new term). We are developing and learning how to co-exist with new kinds of learning entities, the process that rhymes with biology, but is fundamentally different from it in the underlying substrate (what Deleuze would call "risome").

Mossing and others, both at OpenAI and at rival firms including Anthropic and Google DeepMind, are ... studying them [LLMs] as if they were doing biology or neuroscience on vast living creatures—city-size xenomorphs that have appeared in our midst.

Anthropic and others have developed tools to let them trace certain paths that activations follow, revealing mechanisms and pathways inside a model much as a brain scan can reveal patterns of activity inside a brain. Such an approach to studying the internal workings of a model is known as mechanistic interpretability. “This is very much a biological type of analysis,” says Batson. “It’s not like math or physics.”

Anthropic invented a way to make large language models easier to understand by building a special second model (using a type of neural network called a sparse autoencoder) that works in a more transparent way than normal LLMs. This second model is then trained to mimic the behavior of the model the researchers want to study.

Creating a model that behaves in predictable ways in specific scenarios requires making assumptions about what the inner state of that model might be in those scenarios. But that only works if large language models have something analogous to the mental coherence that most people do.

And that might not be the case.

...
Another possible solution ... Instead of relying on imperfect techniques for insight into what they’re doing, why not build an LLM that’s easier to understand in the first place?

https://www.technologyreview.com/2026/01/12/1129782/ai-large-language-models-biology-alien-autopsy/

The biological complexity issue is tricky because we don't want to confuse the complexity of structure with the complexity of behavior. For example, my dog is an extremely complex biological system, but getting/training her to sit is not a big deal. But as we crank up the complexity of behavior, our ability to understand and predict outcomes goes down dramatically.

Because the epistemic value of drawing in science is generally understood in terms of the drawing as an object and not, as is argued at various points in this volume, in the drawing process itself, the unique epistemological value of drawing has become increasingly under-appreciated (Anderson 2014; Tversky 2010; Wittmann 2011). Educationally, this has had costs in terms of student engagement, a deeper understanding of the conventions of scientific representations, and the development of ‘creative reasoning’ (Ainsworth et al. 2011).... A unique advantage of drawing is its double nature as a tool for deepening understanding and for communication. Drawing helps to consolidate ideas, clarify concepts and bring visibility to thought. Drawing can extract and highlight salient information from what is discussed, observed and witnessed, and can be used to communicate and analyze an idea or a concept within a research team or to the broader public.

Drawing Processes of Life, 2024.

Contrary to the common belief that moving images help us to see change, we contend that stillness that helps us to see movement and change. Change is constituted through the relationality of the picture series. The series is a synthesis and analysis of development at once. Development is both the individual form and the series of forms – it is stasis as much as flow. locating development in the images we have created is to imagine development as a pictorial relationship. It is the relationship between the visual forms that produce both the individual stage of development and development as a whole – simultaneously and in mutual dependence (Wellmann 2017: 233).
...
The ‘fishness’ of the fish is maintained through the transformations; each stage bears a unique gradation of colour, repetition of the same form seen from different angles and scales, moving forwards, backwards and sideways, creating a feeling of emergence from a dark background. Klee’s watercol- our washes elevate the forms to a resonant poetry.
...
The essence of rhythm is the fusion of sameness and novelty; so that the whole never loses the essential unity of the pattern, while the parts exhibit the contrast arising from the novelty of their detail’ (Whitehead in Wellmann 2017: 25).

Quoted from Drawing Processes of Life, 2024

Simultaneity of images seems to be key for this representation. Current AR tech is going that way, which might help develop a new way to describe life processes and in general adopt process philosophy-based view of life. Generative AI might help too.

Following our previous collaborations, it had been our intention from the start to make a connected image that would permit us to visual- ize the previously disconnected sequence of changes in topology all at once. This is completely unlike making an animated movie from a sequence of stills, because the movie still unfolds in time, whereas our image is timeless. ‘The whole is no longer subjected to time but rather possesses time within itself’ (Wellmann 2017: 83).

Quoted from Drawing Processes of Life, 2024.

Here, the intent to include time into the whole is articulated so succinctly that I'm going to steal it in its entirety. From now on, all categories should include time by default (see e.g. Lawvere's take on Hegel.)

...the hierarchy of praxis that Ricoeur presents are (1) practices, (2) life plans and (3) narrative unity of life.

--- Wessel Reijers, Mark Coeckelbergh. Narrative and Technology Ethics.

So far, the AI community, esp. its business leaders, has failed to address these three essential narratives of praxis for people outside of the community itself. If you are a truck driver, there's nothing for you in the AI-powered future. By contrast, the web, the iPhone and even social networks accomplished the task very well.

A good checklist:

“... modules are task specific, problem specific, or opportunity specific as often as domain specific, if not more often. Still, ontology is a terrain that inferential modules typically exploit.”

--Dan Sperber. “The Enigma of Reason: A New Theory of Human Understanding.”

I guess, "task" here is synonymous to "routine."

(ix) That how an actual entity becomes constitutes what that actual entity is;+ so that the two descriptions of an actual entity are not inde- pendent. Its 'being' is [35] constituted by its 'becoming.' This is the 'prin- ciple of process.'

--- Whitehead, Process and Reality. (Categories of Explanation).

When you put your money in the bank you (your money) provide(s) a service to the bank, not the other way around like the vast majority of people tend to think.

Similarly, investors serve startups, which explains why hype cycle is similar to Turchin's elite-peasants exploitation pattern.

...the creative advance of the world is the becoming, the perishing, and the objective immortalities of those things which jointly constitute stubborn fact.
...
There remains the final reflection, how shallow, puny, and imperfect are efforts to sound the depths in the nature of things. In philosophical discussion, the merest hint of dogmatic certainty as to finality of statement is an exhibition of folly.

--- A.N. Whitehead, Process and Reality, 1979, 2nd ed.

( Read more... )

We are getting to a point where the Senate sets a precedent that the President of the United States is officially above the law and the impeachment process as defined by the Constitution doesn't work because the President can simply ignore inquiries (Contempt of Congress) as long as they have protection of their political party. Legal remedies are available only after the President leaves the office, subject to the statue of limitations.

upd. The President can effectively pardon themself and avoid any prosecution for federal crimes even after leaving the office.

Based on my brief exposure to CT, in particular to Coecke's work and Milewski's inspirational messages, one of the topics in the summer seminar will be the Structure of Time. Some of the weird things to explain to the lay audience would be why in CT a sequence of actions is recorded in reverse order. For example, if I perform actions x, then y, then z, in CT it's denoted as z○y○x. BTW, Mac doesn't have ○ in its bag of mathematical symbols.

In one view it was the gold and silver per se that nurtured that capitalism. In fact, it was not the metal but its consequences, and these were not at all mysterious. The high prices and low wages meant high profits. From the high profits came high savings and a strong incentive to their investment.

p. 16

note the relative ease of control for different flows.