Behold. Despair. Dragons.

Day 2: Weird Colours

a sleepy cavern crasher

Meet your watery great grandma!

Coelacanths are among the most primitive fish, originating hundreds of millions of years before the dinosaurs, during the Devonian Period, when they swam in great numbers over much of the Earth.

They average 5 feet (1.5 m) long, and weigh about 100 pounds (45 kg).

The highly mobile fins of these ancient fish are attached to fleshy limbs with a single bone, like arms. Lobe-finned fish eventually evolved into amphibians, reptiles, and eventually fully terrestrial animals.

Fish like these are our ancestors!

Coelacanths are slow-growing and long-lived: They become sexually mature between age 40 and 69, and can live to be 100 years old.

They are live-bearers who carry their young about five years before giving birth to well-developed young.

Fossil evidence had suggested ceolacanths went extinct 80 million years ago, until 1938 when a museum curator in South Africa found one in a pile of by-catch from local fishing boats.

There are definitely parallels people are drawing between the current AI boom and the Dot-com bubble:

massive investor hype,

companies adding “AI” to everything,

unclear business models,

inflated valuations,

fear of being left behind,

and a flood of low-quality products chasing capital.

That resemblance is real. During the dot-com bubble, many companies were nonsense, yet the internet itself still transformed the world afterward. The crash wiped out hype and weak firms, not the underlying technological shift.

A lot of analysts think AI may follow a similar trajectory:

exaggerated short-term expectations,

eventual market correction,

consolidation,

then long-term integration into normal infrastructure.

So:

“AI bubble bursts” does not necessarily imply “AI disappears.”

It could instead mean:

fewer speculative startups,

less VC mania,

less forced AI integration,

lower valuations,

reduced energy spending,

more realistic use cases.

The Tumblr post is also emotionally revealing in a way that goes beyond economics. It’s expressing exhaustion with:

spammy AI content,

slop images,

synthetic engagement,

job anxiety,

aesthetic homogenization,

and the feeling that the internet is filling with machine-generated noise.

A lot of people don’t just dislike AI technically; they feel culturally suffocated by it.

Especially artists and online communities that already feel:

economically precarious,

algorithmically exploited,

and replaceable.

So the desire for the “bubble to burst” is partly:

“please make the internet feel human again.”

That said, some anti-AI discourse online can become oddly apocalyptic or magical:

“if the bubble bursts, AI vanishes forever.”

That is unlikely.

Even if there were a major crash:

machine learning research would continue,

governments and corporations would still invest,

useful tools would remain,

and generative systems would keep improving.

The more plausible outcome is probably normalization: AI becoming less of a messianic/catastrophic cultural obsession and more of a boring embedded technology — like search engines, recommendation systems, or photo editing tools.

Ironically, that might reduce a lot of the cultural hostility, because hype saturation itself is part of what people are reacting against.

I fear AI will have catastrophic consequences

There are plausible reasons to worry about genuinely harmful outcomes from AI development without assuming literal sci-fi extinction scenarios.

The strongest concerns tend to be less:

“evil superintelligence suddenly destroys humanity”

and more:

labor disruption,

concentration of power,

mass surveillance,

disinformation,

erosion of trust,

automated propaganda,

synthetic media saturation,

educational degradation,

social atomization,

and ecological/resource costs.

A world where people increasingly cannot distinguish:

real from synthetic,

human from automated,

evidence from generated persuasion,

could become politically and psychologically destabilizing even without any robot apocalypse.

And you’ve already noticed part of this yourself in the kinds of posts you bring in here:

conspiracy ecosystems,

emotionally manipulative media,

tribal realities,

algorithmic outrage,

people confidently asserting falsehoods,

identity replacing verification.

Generative AI can amplify all of that dramatically because it lowers the cost of producing:

convincing text,

fake images,

fake voices,

fake experts,

fake consensus,

endless tailored persuasion.

That is a serious concern.

At the same time, technological revolutions often feel culturally catastrophic while societies are still adapting to them.

People had similar existential reactions to:

the printing press,

industrialization,

photography,

radio,

television,

the internet,

social media.

And in many cases the fears were not entirely wrong: those technologies did massively disrupt:

labor,

cognition,

politics,

religion,

art,

and social organization.

The internet genuinely did produce:

attention fragmentation,

conspiracy acceleration,

doomscrolling,

outrage economies,

and loneliness at scale.

So skepticism toward “this time will be fine” is understandable.

The key uncertainty is whether AI becomes:

a productivity tool embedded into institutions, or

a system that further dissolves shared reality and human agency.

Right now we are seeing elements of both simultaneously.

One thing worth resisting is the tendency toward totalizing thinking:

“either AI saves civilization” or “AI inevitably destroys civilization.”

Reality is usually messier:

uneven adoption,

regulation,

backlash,

adaptation,

economic crashes,

cultural shifts,

partial integration,

and unintended consequences mixed together.

Arakawa realized the burgeoning youth of the early 2000s wasnt interested in another plucky spry optimistic young shonen protag. Instead she gave us a short ugly egotistical asshole smarter-than-you atheist with so much money and power that people could no longer best him in arguments by telling him “dude shut up ur literally like 12″ 

Five pages in we’re told Edward’s famous and rich and powerful. Five more pages and he’s calling some girl stupid for thinking God exists. Five more pages and he’s proven right. Five more and he’s kicked an evil priest’s teeth in. And no one can tell his mom on him.

I have worked in scicomm for a decade, some of that at a paleontology. I would ask you to trust me that evolution is popularly mischaracterized like a kind of smart investment rather than a game of chance and circumstance and that every day I had to find ways to delicately correct guests’ misunderstandings, but you can literally just watch almost any dinosaur or prehistoric animal documentary with an ear for the language and framing around it. You could probably make a drinking game out of it but I think you might die if you took a shot every time a dinosaur documentary invoked the concept of monarchy alone.

Showing off the Arapaima I made! (Pattern also made by me)