Let’s be Fwends #161:

Nuance

Its official title was Constructive Historical Integrative Analysis. In the end, though, there was another name the Minds used, amongst themselves, for this technique, which was Just Guessing.

~ Ian M Banks, The Hydrogen Sonata

How Nuance is Lost (and How it is Regained)

The world is loud. Maybe it was always this loud. Maybe this is new. But if you don’t speak up in a video call, offer an insight, a solution and an action to a problem nobody really understands at the moment, you’re drowned out by the many people who do.

I’ve seen so many knee-jerk reactions advertised that pretend to solve an issue but actually do nothing because no-one understands the issue yet.

If you think about things in a quiet and balanced way, you add nuance to the conversation. Or rather: Would. Because people like clear solutions, presented loudly. They have a much harder time with anything that even remotely sounds like “I don’t know yet, but neither do you.”

How can you improve the chances your position will be heard anyways? By being assertive and purposeful. This is how can bring nuance to a conversation that is otherwise dominated by black-and-white thinking and solutioning.

Analysis-Paralysis

Here’s an interesting thought: When making hard choices, don’t think too much about it.

“The decision-making business is really just the regret-minimization business. And because my big thing is about getting people to make decisions faster, we just have to accept that yes, occasionally, we will regret our choice. We can eliminate a lot of regret by just moving faster. Many more people regret taking too long to make a decision than making a decision too quickly.”

(Source)

There’s a lot to say about making decisions, and the advice to don’t overthink them is solid.

Personally, I’m a big fan of this very basic categorisation of decisions and their consequences:

If they’re reversible and inconsequential, decide on the spot. If they’re consequential, review existing data fist. But if they’re irreversible, and consequential, do lots of research up front, and start thinking in scenarios.

(If they’re irreversible and inconsequential, you also decide on the spot after reviewing data)

When you honestly perform this categorisation first, you’ll notice how many decisions are either reversible or relatively inconsequential. Dealing with them quickly allows more mental space for dealing with the decisions that deserve your full attention.

Cognitive Debt When Using AI

Speaking of “thinking less” - we need to add a new concept to our vocabulary: Cognitive Debt.

In an essay-writing experiment, researchers divided participants into three groups: A group that used an LLM for writing, one with access to a search engine, and one that was allowed no tools.

What they found was quite disturbing:

“EEG revealed significant differences in brain connectivity: Brain-only participants exhibited the strongest, most distributed networks; Search Engine users showed moderate engagement; and LLM users displayed the weakest connectivity. Cognitive activity scaled down in relation to external tool use.”

(Source)

It’s obvious that people who used an AI assistant to generate essays would have trouble identifying with the results, feel a sense of ownership or citing “themselves”. But the fact that one of the reasons for performing mental tasks in the first place - the training and enhancement of the organ who does all the thinking is significantly diminished with tool usage should make us all pause.

But who am I kidding, of course if’s “full steam ahead, no time to think twice” on the AI ship.

Going Dark

OpenAI published their latest report on malicious uses of AI, and it is definitely worth a read.

Bruce Schneier points out the real danger here:

“Reports like these give a brief window into the ways AI is being used by malicious actors around the world. I say “brief” because last year the models weren’t good enough for these sorts of things, and next year the threat actors will run their AI models locally—and we won’t have this kind of visibility.”

(Source)

Understanding of (and ultimately, defence against) AI-enhanced attacks will be extremely difficult if we only see the result of the attack construction, and not the complete design.

A BIC Success

Sorry, I could not resist that bad pun. The BIC ballpoint pen is one of the greatest writing utensils ever invented (both the clear and the orange ones). It’s also arguably the most successful product in history, selling over 100 billion units.

It also is a masterpiece of design and technology, with a process of creating high-precision stainless steel balls at it’s center.

Changing Scale, the Analog Way

Do you know what a pantograph is? I have to admit, I did not. It is a device with an incredibly simple construction that allows you to scale anything you write or draw.

You Didn’t Do It and That’s OK

If you’re like me, you have tons of unfinished projects. And if you have trouble to let go of them, try one of these Unfinished Projects Release Forms - “a permission slip for emotionally letting go of a project you won’t finish.”.

Let’s Be Fwends From the Past

Four years ago, I shared a report about the number of hours you should work in a day, and now that summer holidays are approaching, this number is going down significantly.

A year ago, I’ve been writing about AI and possible consequences for your personal growth. And I don’t want to turn this into a “I told you so” moment, but - well: I told you so.

That’s it for this edition of Let’s be fiends - thanks for reading, in any scale you want. 📏