Nuance in the Internet era

The internet, especially social media, appears divided into two camps: pro-AI and anti-AI.

Pro-AI people see AI as the next frontier, and anti-AI people as uncultured luddites.

Anti-AI people see AI as a dangerous tool of oppression that is accelerating inequality and climate change, and pro-AI people as fascists.

Linus Torvalds in 2024 said:

I think the whole tech industry around AI is in a very bad position and it’s 90 percent marketing and ten percent reality.

Linus Torvalds

Then, in 2026, he commented:

AI is a tool, just like other tools we use. And it’s clearly a useful one.

It may not have been that “clearly” even just a year ago, but it’s no longer in question today.

Linus Torvalds

Linus Torvalds was never an anti-AI person, and he was ready to change his opinion as soon as the technology matured and proved useful.

Daniel Stenberg also has a nuanced view on LLMs. He went from removing payouts for curl on HackerOne in January 2026 due to slop reports to allowing Mythos to scan the curl codebase in May 2026, which he admitted that, despite Mythos being overhyped and it only finding a single vulnerability, was “still, pretty good”.

Uncle Bob, the author of “Clean Code”, said:

My current strategy is to not read any of the code written by my agents. That’s the only way I can take advantage of their productivity. What I do instead is to surround the agents with extreme constraints. Unit tests, gherkin tests, QA procedures, quality metrics, mutation testing, test coverage, and a plethora of others. In the end, I have very high confidence in the code they produce because they’ve had to run the gauntlet of all of my constraints and tests.

Uncle Bob

Even John Carmack has a positive attitude towards AI. He’s credited with the rendering engines of Quake, Quake II, Wolfenstein 3D and Doom.

It seems to me like a few high-profile individuals are embracing AI and approaching it pragmatically.

Say what you want about AI, but these highly successful, very well-respected software engineers and programmers clearly believe AI is useful to some degree.

On the other side, there’s plenty of criticism against AI coming from experienced engineers:

• Krita Lead Developer doesn’t want LLM-generated code in Krita
• Codeberg no longer allows LLM projects
• Drew DeVault denounces LLMs
• Asahi Linux prohibits “slop generators”
• QEMU declines any contributions which are believed to be AI-generated content [archived version]

For some of these anti-AI stances, I get it! I understand what DeVault says about the ethics and the (very real) cost of having to serve traffic to AI crawlers that often use residential proxies to evade anti-bot mechanisms. I also understand the concern some people have with the licensing or legal status of AI-generated code. I think some of these points are valid and worthy of consideration.

Where’s the nuance?

There are very few absolute certainties in life, but it would take you a few minutes of browsing Twitter to see how certain people are regarding their own beliefs. They’re so certain, in fact, that they will fight and argue with strangers on the Internet to defend their own ideas — for free!

The current AI discourse is but one example of how polarised the Internet commentary has become.

It is hard to argue against the fact that the Internet has brought everyone closer together. The number of people you can meet and communicate with has exploded, and these days it is only a matter of timezone coordination.

However, while the Internet has made communication faster, easier, and more frictionless, it has also encouraged and facilitated hostility.

Nicholas Carr explores this paradox in his book Superbloom, which I strongly encourage you to read.

Some of my favourite quotes from this book, in no particular order:

With an almost microscopic view of what everybody else is saying and doing […] we have no end of opportunities to take offense.

Nicholas Carr, Superbloom

We have a bias to like people that seem similar to us and dislike those who seem different.

Nicholas Carr, Superbloom

Sherry Turkle, an MIT social psychologist who has been studying how people communicate through computers for decades, has described social media as an “anti-empathy machine”. She argues that we suppress our capacity for empathy by “putting ourselves in environments where we’re not looking at each other in the eye, not sticking with the other person long enough or hard enough to follow what they’re feeling”.

Nicholas Carr, Superbloom

The book also goes on to explain how we¹ thought that, by making communication more instantaneous, less lossy, frictionless, and affordable, we would solve many world issues, including the appropriate allocation of resources, and we might even permanently end wars because there would be space to debate and discuss, and there wouldn’t be misunderstandings.

Sadly, it seems there are unresolvable ideological and political differences that separate us.

Of course, it is not just AI; it is almost any topic you can imagine. Climate, abortion, masculinity, LGBTQ, renewable energy, religion (and freedom of religion)… There are only two camps, and if you are not unreservedly in either camp, then you’re in neither.

And of course I understand why. The climate is being destroyed piece by piece. Every single scientific prediction is slowly coming true. Scientists gave us warnings; we didn’t do enough, and what they said would happen is now happening. Women are being stripped of their right to abortion. Governments worldwide are cracking down on “LGBTQ content”. And so on. It’s hard to remain neutral about any of these topics when society is being dismantled piece by piece.

I think it is a total waste of potential and resources that we have these vast interconnected social networks and complex computer systems, and they’re mostly used to feed our own biases back to us, keep us hooked, and agitate us in order to maximise user retention and show as many ads as possible.

Conclusion

I don’t think the lack of nuance in online discourse is good for anyone or anything. I also don’t think this will ever change.

It’s not what I want, but it is what it is.

But I suspect I may not be alone in feeling this way. I know for a fact some people hold very sophisticated and nuanced opinions, because the world is not black or white. And I know they are afraid to express them, because these opinions, while rational, are highly unpopular.

We² probably need to be building more online spaces that provide people with space to pause, reason, reflect, and pause again. Maybe we need actually more friction, and not less? But I don’t know how to get started. It’s definitely not something I can achieve alone.