I don't think we can say what will happen, but I agree that history offers no guarantees for the future.
3,118 karma · joined January 28, 2016
The frontier labs are investing TONS of capital into training the new 'best model ever'. Today, that doesn't make economic sense, it loses money. So where's the business? The future, where token costs reflect the costs involved.
For that to be corrected, the labs need to get people using the models for things they depend on, then they turn dial up on cost. The ordering there is critical. It has to become indispensable first, then costs go up.
The trouble is, open weight models are rapidly becoming usable in what is looking like the big market: coding. Step 2 (the costs go up) doesn't work if there is a ready off-ramp that doesn't require frontier models.
However, if they can effectively make open weight model development illegal (its dangerous!), their business strategy can remain intact. A nice side effect is that they will be effectively be granted an oligopoly "for humanity's safety".
If they can convince the populace that a text generator is an existential threat to humanity, that's a hard argument to combat.
Edit: I think it is ridiculous that autocomplete on steroids is an existential threat. There's no way anything harmful happens until you actually use it as part of important systems. By that reasoning, you could make it illegal to use AI in important systems, but do you think that's the argument the frontier labs are making to governments?
By arguing that the AI is dangerous, they are really arguing that someone 'trusted' needs to monitor the situation, so that the outcome is aligned with existing power.
They are trying to scare the world into forbidding others the right to use and develop AI.
The 'that was for my assistant' is just something they say to explain away the earlier quoted price. They are pretending that what they said wasn't the quote for this conversation, it was for another, different conversation, between the salesman and their assistant. This only works when the assistant is not in the call.
https://sustainability.google/reports/google-2026-environmen...
https://www.aboutamazon.com/news/sustainability/amazon-data-...
I don't think 'anyone who thinks X is Y' is a good basis for discussion, which is the real point.
> Approximately 80% of the water (typically freshwater) withdrawn by data centers evaporates, with the remaining water discharged to municipal wastewater facilities.
https://www.eesi.org/articles/view/data-centers-and-water-co...
According to EESI (a bipartisan-founded non-profit), a large data center:
> can consume up to 5 million gallons per day, equivalent to the water use of a town populated by 10,000 to 50,000 people.
Also noting that AI datacenters can use more.
Filter and map are easy because you can make more assertions about them without inspection:
- they take a list-like as input
- they take a function as input
- they return a list-like as output
As opposed to reduce:
- it takes a list-like as input
- it takes a function as input
The output type is not fixed, and its behavior is not fixed. If you pass the right function, reduce is filter, or map, or something else we've never seen.
I might accept a Meta paperclip into my house, but then again, why risk it ;)
Meta's game is to do all sorts of dirty tricks to pull you into its data. No thanks.
No level of convenience will tip that scale.
I'm sorry for those who don't have people in their lives, but machines aren't a substitute, and that lack shouldn't be used to justify a right to unbounded freedom to experiment without regard for the possible negative consequences.
If I could run linux on Apple hardware and get the same power-sipping performance, that's what I'd be doing.
I move much faster when I don't have to take my hands off the keys to nudge the rat.
However, I don't think my preference for TUIs is about my efficiency. The most important reason I love TUIs is that they generally only have a base set of features needed to get the job done. Most GUIs go off the rails, implementing features for use cases few people have. TUIs tend to be small, fast, and focus on a small use case.
So even though I started out claiming I love TUIs, what I really love is small software: CLI > TUI > GUI.
>> Similarly to how you can just eat less and lose weight.
I'd argue it isn't similar in an important way, people cannot opt out of eating entirely. Overeaters must eat. People who can't control themselves with social media do not need to use them at all.
His criticism wasn't for or against TigerBeetle, it was against the religious 'mode of persuasion'.
`The faith isn’t needed. We need to let the work stand on its merits — show the receipts, empower people with the facts and reason, and trust them to make the right calls.`
Seems like that would add significant cost for something that can be done well without it.
There are groups of people using the same terms for different things, and getting into arguments where there could be agreement.
There's a diverse set of people on the spectrum between formally trained and self taught people in software development. Terms can get borrowed from places like academic research and get (mis)applied before they are understood.
The REST in industry software is essentially JSON RPC (no, not that one) mapped to HTTP methods. That's a mouthful, so people just say RESTful, despite it not quite being RESTful according to the original notion of REST.
Yes, industry-RESTful isn't necessarily RESTful according to original Roy Fielding definition, but trying to redefine a widely used industry term is unlikely to win one friends at dinner parties.
In other words, this probably won't directly involve changes to ghostty.
The fascination with LLM coding seems rooted in the new ability to create something with very little effort. Unfortunately, some people do not see the trade comes at the moment you have to evaluate the output for fitness or quality.
It turns out that you often need to understand the thing pretty well in order to be qualified to determine the quality of the output. That tends to dilute time gains, unless you leave off the evaluation step.
Unprofessional? Maybe a little, but honest, and while the truth isn't flattering to Jarred, I'd say generally kind.
Embarrassing? Not from my seat. Andrew is just revealing the relationship dynamics between a principled programming language developers and pragmatic business users of that language.
I had already inferred a bunch that this post confirms based on the agentic port from forked-Zig to Rust. It's very nice to have suspicions confirmed.
I honestly don't think windows is worth any price. Unix is the platform of builders, and nobody needs a middleman charging them to breathe air (run an os).
For desktop use, sure, $45 could fly, but, the main draw of windows is compatibility with existing software, and honestly that is no longer worth it.