Haha fair I just like to ask basic questions to listen more sometimes

I can totally understand that. Being a manager is incredibly exhausting and I've always wished there were more IC technical tracks available. It is basically not an option for research. Hope you get a good environment that allows lots of cognitive ease

You would like this talk I have been drafting about what deep learning in agentic coding could look like ... Hope to get it to a state to do a public version

as I get back to a lot of hands-on solo work!

I try to distinguish btw my deep structural understanding of a domain area & find ways to keep building it across different context-specific implementation of tool skills or process skills. I'm less of a fluent R coder now but have more expertise in research design. I can see that shifting back too

I think all of us face a lot of skill change across our lives if we persist long enough in a career; I feel needing to learn code takes me away from learning scientific skills that I value! But affords very different contributions to the world, as becoming a leader afforded me.

Is what bad?

I'm so glad you're experiencing such a positive effect! I can't say without knowing more but I feel finding structures that work well for our minds can be incredibly helpful and beneficial. Does it feel really different to how you wrote software before?

allowed!!!

Yes there are the kind of claims where if it's not specified more precisely it's impossible to speak accurately about the relationships. Effort is good and also too much effort will kill you and also not enough effort will kill you. Absurd to scaremonger about this (bad) preprint

Separate from THAT methodological problem, if you can induce a relatively sharp decrease in a measure that you can also easily induce by other everyday experiences (e.g., typical states) that's mostly an argument against structure change not for it. Don't even need to know EEG to reason that out

Separate from the utter drek of making this pseudoscience chart based on nothing, I don't trust EEG functional connectivity estimates in the first place and neither should you.

Chart of the Day: How Using LLMs is Like Driving Drunk There is a growing body of work on the effects of large language models on the brain, from cognitive offloading, to direct measures of brain activity. While all of this work is interesting, the latter...

I've seen some wild pseudoscience charts in my time but WOW this one killed me. Brain astrology

paulkedrosky.com/chart-of-the...

There is indeed a whole world out here. Lol some of these takes reveal they've never been in it

Reading these threads and quotes as a PhD in industry who has created a research dept with PhDs is always like sitting in a movie theater watching an old timey film

Totally, states are environmentally responsive (we often assess the impact of different environments with states as an outcome)

Oh never would I compare the magnitude of this

Yep -- that recalibration is EXTREMELY important for complex knowledge work jobs!

😎

The questions become a lot more answerable, when you force yourself to stop reasoning in generalities and really look at real behaviors and real states as unfolding over time - I have really found it to be true not just for scientific curiosity but for landing on things that help others.

Suddenly we can ask different questions. What *states* do you want to be able to move in and out of? Who is allowed to do that? Can someone use a different tool to access the same state? Even if we treasure a certain way of encouraging thinking, might there be others?

In fact, let us be free for the moment from roles at all. Those are words on some HR dashboard but thinking is a thing you get to do. Think about states instead, which is what psychologists of change like me like to think about.

I don't like the idea of mapping these things onto professional roles and reaffirming already deeply-held stereotypes about who is a good thinker or who can do what, vs thinking deeply about the real types of behaviors we are talking about and their very different forms of expression

The point of this thought exercise is that I'm starting to see a lot of patterns in the research literature around AI that are essentially about whether people are motivated to take agency and active stances toward the tool(s). These are psychological stances and we know things about them

I can imagine someone who never learned to code who is nonetheless highly active & adaptive with technology; curious, personalizes things. My grandfather would've fit this, constantly manipulating the environment around him, taking agency over figuring out how and why things worked.

I can imagine a developer (blameless description, let's say there are lots of environmental reasons this is true, state not trait!) who is copy-pasting solutions, manually writing code, and has signifiers we associate with "deep thinking" but really stays at the level of passive orientation

I can imagine some stereotypical technology users we could map onto this. Perhaps we think ok software developers are "active" and non-technical users are "passive" toward technology.

But I also like to play around with thinking well, can I imagine a "passive orientation" developer? I sure can

Think about some features that might mark an "active" orientation toward technology. I can imagine verifying others' opinions via self-testing, pursuing mastery knowledge, experiment-based understanding, high experimentation and personalization (again, speculative)

Think about some features that might mark a "passive" orientation toward technology. I can imagine taking others' opinions, surface level knowledge, consumptive understanding, low experimentation and high receptivity/reactiveness to changes. These are all just speculative

Let's imagine you could measure a passive vs active orientation toward technology as a construct *

*ie let's pretend we've done a lot of nice psychometric work to develop some scale that reflects this and we feel good about it and we share a philosophical belief that social science is useful