Post · bonfire.cafe

@grimalkina@mastodon.social · 11 hours ago

RE: https://mastodon.social/@jilleduffy/116593522913243021

Setting aside whether one thinks AI is successfully automating, I disagree with this argument because automating "the mindless parts of work" has generally been really good for people. Look at studies of factory workers: those tasks are murder on your executive functioning, are very draining, and you have to be vigilant to do highly repetitive "mindless" work. It is really terrible for people and generally speaking we like complex knowledge work far more

I agree about taking breaks being good

jilleduffy

@jilleduffy@mastodon.social · 11 hours ago

You know what else about AI?

People who hype it have no idea about diminishing returns

We have decades of research showing that when people only work on the hard stuff at work and don't get sufficient breaks, they burn out faster

If you're selling AI as a tool that "automates the mindless parts of work" then the worker is only left with the hard parts, and they burn out faster

Sue Smith

@sue@glasgow.social · 5 hours ago

@grimalkina Aye, class related blindspots aside, we need people to reflect on the detail and nature of the different types of work they do to make good decisions about what to automate (in their context). To flip this on its head slightly, one of the prompts I've been using is asking folk to consider where there is value in the process rather than just in the output.. Where there isn't, automate away. It's not a simple task and likely needs revisited as a practice.

Michael Newton

@mavnn@bonfire.mavnn.eu · 2 hours ago

@sue@glasgow.social @grimalkina@mastodon.social

> there is value in the process rather than just in the output

This seems a crucial aspect that is frequently being ignored, and I suppose I see this highlighted as somebody who's doing some teaching at the moment. Brandon Sanderson (the author) has a saying he's given several variations of that boils down to "I didn't get any money for all the books I wrote before I got published, but I did get a better writer and a better person from them."

On the other hand, I started programming by literally automating my then data entry job out of existance. I didn't learn any more manually copying the data between a spreadsheet and a database the third time than the second.

But I think we're terrible at judging what is and what isn't a learning experience and worth doing, and even worse I think that can change with both experience and situation. My advice on whether to use an automatic website creator is going to different if you're an art student who needs an online portfolio right now, or if you want to be a website designer.

Sue Smith

@sue@glasgow.social · 2 hours ago

@mavnn @grimalkina absolutely, it's extremely contextual which is why I see it as a practice or perhaps framework we can develop and iterate on

Michael Newton

@mavnn@bonfire.mavnn.eu · 2 hours ago

@sue@glasgow.social @grimalkina@mastodon.social Of course, I have some other concerns around LLMs as 'automation' given the relative failure rates compared to other things I would call automation (plus ethics, legality, etc). But the 'can it be done' discussions and the 'should we use this tool' discussions seem to get more time than the 'is this even a good goal' discussions.

jilleduffy

@jilleduffy@mastodon.social · 11 hours ago

@grimalkina Thank you. I think probably I should have made a distinction between automating factory work versus knowledge work, although you still have a point that there is some amount of automation that is of course useful. Templates are useful, for example.

The "hype" to me around AI says "figure out how to automate as much as you can with AI or you will be left behind," and I just don't think that takes into perspective how most knowledge workers actually work

Cat Hicks

@grimalkina@mastodon.social · 11 hours ago

@jilleduffy I definitely strongly agree we need to pay careful attention to how people work and problem solve and what they need!! For any proposed technology and especially this one

Cat Hicks

@grimalkina@mastodon.social · 11 hours ago

@jilleduffy I feel like I understand what you are saying when I think about the benefits of mind wandering, and more reflective tasks, there is a certain ease to it all when you get to have a somewhat low effort/high flow feeling task around your deep focus work!

Claudius Link

@realn2s@infosec.exchange · 10 hours ago

@grimalkina @jilleduffy
Thank you for the discussion and impulses.

Additional to all said the "mindless" part is sometimes hard to separate from the "hard" part (especially if the hard part is somewhat time critical)

Examples could be autonomous driving where it's "easy" to automate the standard driving but hard to transfer control to the human in case of a non standard situation

Or train operators who have to press a button in intervals just to keep their attention

Cat Hicks

@grimalkina@mastodon.social · 11 hours ago

A huge amount of cognitive exhaustion research is in fact about workers who could be saved from pretty horrendous daily lives with automation.

Cat Hicks

@grimalkina@mastodon.social · 11 hours ago

This is often something we misconceive of, this is an older piece but appropriate title! "Vigilance requires hard mental work and is stressful"

https://pubmed.ncbi.nlm.nih.gov/18689050/

PubMed

Vigilance requires hard mental work and is stressful - PubMed

This research applies to most human-machine systems that require human monitoring, particularly those involving automated subsystems.