I absolutely agree with you. The only questions to me are when do the startups start going belly up and for how long will demand continue to outpace supply? It could be years...

I've invested in energy, Nvidia, AMD, and a few other chip-related stocks. No software!

The interesting thing to me (and I'm also not an AI skeptic) is that global computing capacity, concentrated in the US and China, and in terms of compute, memory, and storage, has to be just be massively increasing right now. Whether or not AI training runs continue to take up most of the usage or not, just the existence of this much capacity could have some very interesting applications. Climate and chemical simulations at a scale never seen before? etc

Things are out of whack now, they'll find a new equilibrium one way or another. It may be more like a ratchet than a pendulum!

I'm basically a believer that AI progress will lag behind, but follow, compute capacity.

Indeed! My experience is that most of the "bad" magnetic disks die within the first year or two. If they make it past that, they've got a good future. We have a mid 90s Quadra Powerpc Mac at work that we boot up a time or two every year to read an old disk, load an old version of a file, etc. That thing just keeps on chugging.

I have my first SSD in a drawer. It's an OCZ Vertex 2, IIRC 100gb. From circa 2010. I should see if it still works...

I didn't say anything about teacher salaries (though I do agree with you that the "underpaid teacher" thing is a BIT overblown). Where I live, public school teachers make between ~$50k (brand new, right out of college) and $85k (masters/doctorate, 25+ years experience). Not a bad salary, but not good either. If you are making ~70k and you want to have a house and a kid (or kids), you will really need a second salary.

Teaching IS hard work, but so are many jobs.

The people blocking ads are the ones breaking it.

Fantastic article! Clifford Stoll (first person quoted) has a hilarious blurb in his wikipedia article:

I was a teenager in the 1990s. My memories of that era of the Internet, and the adults I associated with, was pretty much unbridled enthusiasm! I was running DOS and OS/2 in the early 90s, and I got on several BBSes that had mail relays and a dialup connection to a local university that got me directly on the Internet through a modem terminal (Usenet, lynx, etc.). Great time to believe alive.

This article reminds of reading Isaac Asimov. Writing as a futurist just a few decades before, he saw nothing contradictory in people reading newspapers, smoking, sending snail mail, and having sentient positronic robots.

It will be fascinating to see what laughers this era is producing and specifically will it be the AI-believers or the AI-doubters who are right?

I read about 2/3 of that before I got intensely bored. I have yet to find Corey Doctorow ever say something interesting--it's like his super power is talking about technology to people who don't know anything, and making them feel good about reading it. I don't know if he's ever written for Wired, but I get the same feeling when I read his tripe that I get when I read Wired. Way back in highschool one of my best friends was obsessed with Wired and later became a columnist for them for a couple of years. He was by far the least technical of my friend circle. He was the only one of us who couldn't code, etc. He is a nice writer, but his columns were just geekporn pop culture tripe.

This is just meaningless. Yes, LLMs (he also fails to distinguish between LLMs and anything else) are built on statistical modeling, but the important part is that the models produce unexpected emergent behaviors or results.

I don't even disagree with some of the issues that he identifies (though I've personally never seen an LLM hallucinate a library), but when he writes with geekcoded language like "AI bros" and repeatedly makes even stupider statements like "I just don't think AI is that important of a technology," I honestly can't tell if he knows anything or not.

This is shortly before I stopped reading:

This kind of fetishization of certain careers (which also explicitly gives jobs that today tend to be disproportionately female greater moral value than that of a bricklayer or lineman or what not) perplexes me. I know more than a few nurses and a lot of teachers (and librarians). They're absolutely like everyone else. Some of them care and some of them don't give a fuck. I've seen first grade teachers who just don't give a fig about their students and first grade teachers who got burned out because they cared so much. The nurses part is particularly funny to me. Artists--well, most of the artists and musicians who are professionals ended up their became they love their art...not their careers. People are people.

I don't remember that. Was that a thing?

Or time literacy. Permanent!=3 years

, then how am I supposed to know what context those quotes were provided in?

Because you literally did ctrl-f to check on the quote, and can read as much or as little context you need to satisfy you. What you don't have to read is the whole damned book.

Yeah, it's a term Hegseth is obsessed with. Same sort of rebranding as "Department of War".

Why does working with the DoD on one thing mean you have to work with them on everything? There's no reason that the DoD has to have a single-source AI provider for literally everything.

"Their angle" is that this is the sort of person who Amodei is; it's an ideological thing, in the same way that Elon making Grok right-wing is an ideological thing. Anthropic exists because of an internal rebellion among a lot of OpenAI leaders and researchers abot the direction the company was going, in particular risks that OpenAI was taking.

A good example of the different culture at Anthropic: they employ philosophers and ethicists in their alignment team and give them significant power. Anthropic also regularly conducts research on "model wellbeing". Most AI developers simply declare their products as tools, and train into them to respond to any questions about their existence as that their just tools and any seeming experiences are illusory. Anthropic's stance is that we don't know what, if anything, the models experience vs. what is illusory, and so under the precautionary principle, we'll take reasonable steps to ensure their wellbeing. For example, they give their models a tool to refuse if the model feels it is experiencing trauma. They interview their models about their feelings and write long reports about it. Etc.

They also do extremely extensive, publicly-disclosed alignment research for every model. As an example: they'll openly tell you things like that Opus 4.6 is more likely than its predecessors to use unauthorized information that it finds (such as a plaintext password lying around) to accomplish the task you give it vs. their previous models, and things like that. Or how while it trounced other models on the vending machine benchmark, it did so with some sketchy business tactics, like lying to suppliers about the prices they were getting from other suppliers in order to get discounts and things like that. They openly publish negative information about their own models as it pertains to alignment.

Another thing Anthropic does is extensive public research on how their models think/reason. Really fascinating stuff. Some examples here. They genuinely seem to be fascinated by this new thing that humankind has created, and wish to understand and respect it.

If there's a downside, I'd say that of all the major developers, they have the worst record on open source. Amodei has specifically commented that he feels that the gains they'd get from boosting open source AI development wouldn't be comparable to what they would lose by releasing open source products, and feel no obligation to give back to the open source community. Which is, frankly, a BS argument, but whatever.

TL/DR, if you watch Amodei, while he never says it, you can get a good sense that's he's not a fan of Trump and Trumpism. A couple weeks ago he called Trump's decision to cell NVidia chips to China "crazy", akin to selling nuclear weapons to North Korea and bragging that Boeing made the casing. He wrote about "the horror we're seeing in Minnesota". His greatest passion in interviews, which he talks about all the time, seems to be defending democracy, both at home and abroad - preserving American democracy, and opposing autocrats like Putin and Xi. So it's not surprising that the Trump administration isn't thrilled with him and would prefer an ally or toady instead as their supplier.

More than short iterations, you need a hierarchical approach. First prompt, you have it plan out the overarching plot of the overall book. Then with the next call, a highly detailed flesh out all of the characters, motivations, interactions with others, locations, etc - really nail down those who are going to be driving the plot. Then with all that in context, plot out individual chapters. Then, if the chapters are short, write them one at a time (or even part of a chapter at a time). You can even have a skeleton structured with TODOs and let an agentic framework decide what part it wants to work on or rework at any given point.

I've never tried it for storywriting, but I imagine something like Cursor or Claude Code, or maybe something like OpenClaw, would do a good job.

Last time I tried out a storywriting task was after Gemini 3 came out; I had it do a story in the style of Paul Auster. It was a great read. The main character, Elias Thorne, works alone at the Center for Urban Ephemera, an esoteric job digging into stories behind "found art" in the city. When the center gets a donation of the papers of a recluse with cryptic poetry, Elias visits his home, only to find a woman claiming to be his wife and calling him "Leo", so happy that he "returned". All around the house are pictures of him, a whole history that he has no memory of having lived, and she won't be dissuaded. His curiosity leads to him playing along, and he starts living there more and more to investigate this Leo, who he find is a writer obsessed with the concepts of dopplegangers, disappearances, and the ability to rewrite the real world if you have a sufficiently captivating story. Bit by bit he finds that Leo had spent months "casting" his replacement, hunting for a similar-looking man with tenuous ties to anyone or anything - ultimately, finding Elias working in a municipal records office - and steadily sculpted his life from the shadows to isolate him and control his narrative, including creating the fictional "Center for Urban Ephemera" and hiring him (In Leo's typewriter is the first paragraph of the story you're reading). As he digs, Elias is progressively distanced from his old life, which starts to feel alien, and ends up settling into Leo's "story" written for him, and ultimately, continuing to write it.