tobyjsullivan

The total probabilities add up to 1. But I’m not following how that relates to the average bits.

Despite summing to 1, the exact values of P(true) and P(false) are dependent on the options which have previously been discounted. Then those variables get multiplied by the amount of information gained by either answer.

Has the author redefined “effort” such that the amount of effort required to carry a boulder up a mountain is, by his own definition, always zero?

> Let me share my slightly unusual definition of “effort”: it’s the felt experience of expending energy beyond what an activity requires

The author defines one “bit” as ruling out half the remaining options.

So a yes might rule out 75% of remaining options (for example) which provides 2 bits of information.

Don’t forget cotton.

234 m3 per tonne, of clean water.

25M tonnes per year.

=> 16M m3 of clean water per day

Edit: convert to comparable units

Likewise. I’ll regret it but I certainly won’t be complaining to the Internet that it did what I told it to (skip permission checks, etc.). It’s a feature, not a bug.

No doubt there’s a lesson in there, but I don’t think it’s a technical one.

Despite a couple forward-looking statements, I didn’t read this as a prediction. It seems more of a subjective/anecdotal assessment of where things are in December 2025. (Yes, with some conjecture about the implications for next year.)

Overall, it echos my experience with Claude Opus 4.5 in particular. We’ve passed a threshold (one of several, no doubt).

JavaScript? HTML and HTTP were the real mistakes.

From that article:

After taking his final photograph, Landsburg removed the roll of film from his camera and placed it in a canister. He buried the camera and the film canister deep in his backpack. Then, he placed the backpack on the seat next to him and covered it with his body.

It was more than just jumping on top of his camera. Sounds like there’s some confidence about his intention.

Many (I’d speculate most) workflows involve moving and referencing data across multiple applications. For example, read from a spreadsheet while writing a notion page, then send a link in Slack. If any one app is down, the task is blocked.

Software development is a rare exception to this. We’re often writing from scratch (same with designers, and some other creatives). But these are definitely the exception compared to the broader workforce.

Same concept applies for any app that’s built on top of multiple third-party vendors (increasingly common for critical dependencies of SaaS)

What do you think Cloudflare’s core business is? Because I think it’s two things:

1. DDoS protection

2. Plug n’ Play DNS and TLS (termination)

Neither of those make sense for self-hosted.

Edit: If it’s unclear, #2 doesn’t make sense because if you self-host, it’s no longer plug n’ play. The existing alternatives already serve that case equally well (even better!).

I’m not sure I share this sentiment.

First, let’s set aside the separate question of whether monopolies are bad. They are not good but that’s not the issue here.

As to architecture:

Cloudflare has had some outages recently. However, what’s their uptime over the longer term? If an individual site took on the infra challenges themselves, would they achieve better? I don’t think so.

But there’s a more interesting argument in favour of the status quo.

Assuming cloudflare’s uptime is above average, outages affecting everything at once is actually better for the average internet user.

It might not be intuitive but think about it.

How many Internet services does someone depend on to accomplish something such as their work over a given hour? Maybe 10 directly, and another 100 indirectly? (Make up your own answer, but it’s probably quite a few).

If everything goes offline for one hour per year at the same time, then a person is blocked and unproductive for an hour per year.

On the other hand, if each service experiences the same hour per year of downtime but at different times, then the person is likely to be blocked for closer to 100 hours per year.

It’s not really bad end user experience that every service uses cloudflare. It’s more-so a question of why is cloudflare’s stability seeming to go downhill?

And that’s a fair question. Because if their reliability is below average, then the value prop evaporates.

Seems like it would also result in capital investors covering more year-to-year tax revenue, which could reduce some pressure on other tax payers.

In theory, capital gains should average out over time. But in practice, I think an increasing amount of wealth is being held and not realized over many decades.

It doesn’t help anyone that a few billion $ of gains will be taxed eventually if that is so far into the future that most citizens alive today will have passed away by then.

The analogy conveys that QA is not inherently necessary to build and ship things that work.

It also paints a picture of a scenario where requiring QA would be more of a red flag than a best practice. It seems a tad silly to imagine a woodworker nailing boards together so they look like a table, then passing to QA to determine if the table is “good enough”, then having QA ship it back with defect reports. But this is exactly what many less-mature teams end up looking like.

You make a good point about unexpected interactions.

I’d argue the question for software isn’t whether QA Bad or QA Good. It’s at what level of complexity does QA become necessary. Most software teams aren’t dealing with all that much complexity (or, more specifically, inherent complexity that can’t be designed away.)

It’s possible for a woodworker to build a table which reliably does not collapse, even without having a third party test it.

It’s equally possible for a different woodworker to build a table which will collapse when deployed in a customer’s dining room.

The difference comes down to which woodworker I’ve hired, and how they’ve been trained.

If you can’t trust a woodworker to ship a table that stands under its own weight, layering on third-party QA isn’t really going to fix the underlying problem.

That said, cargo culting the “no QA” model is ill-advised. If a particular dev shop needs QA today, they’ll probably need it tomorrow.

> Google says they charge $1.60 to $2.60 a mile

That’s surprising. I’ve been trying to find data on rates and crowd-sourced data and anecdotes seem closer to $6/mile

This reminds me of the debate around Soylent when that came out. Are meals for enjoyment, flavour, and the experience or are they about consuming nutrients and providing energy?

I’d say that debate was largely philosophical with proponents on both sides. And really the answer might be that both things are true for different people at different times. Though I also observe that soylent did not, by and large, end up replacing meals for the vast majority.

Not exactly this but found a similar experience recently by throwing Linux lite on an 15-year-old thinkpad with a physical wifi switch.

Was for my kid in this case. Loaded a few education-friendly games and then disabled the wifi. Now it’s a simple, focused, and relatively safe box.

Could easily do the same for writing or any other activity.

Rather than too sensitive, I think you’re making up AI signals. Poor writing (or, in this case, slightly less than perfect writing) is not a phenomenon to which humans are immune.

If you like the idea of human-written content on the Internet, I recommend against joining the chorus of voices baselessly accusing humans of being AI bots - an unfortunate trend lately which only serves to disincentivize future contributions.

ChatGPT didn't have any issue when I recently asked something very similar.

https://chatgpt.com/share/6909b7d2-20bc-8011-95b6-8a36f332ac...

Hypothetically, how would someone know there was a warrant out for their arrest in another country? That doesn’t seem like public information.

I figure most cyber criminals assume they are untraceable until they get arrested.

Everything I’ve seen suggests that Microsoft has entered the metaphorical private equity phase of investment in Windows. They’ve already given up any expectation of it being a viable competitor long-term and are purely focused on milking as much short-term revenue from the product as possible before it dies.

I’m sure windows will continue to exist and maybe be relevant for at least a decade. But it will be in zombie/revenue-extraction mode from here on.

It’ll help keep PRs short

How does this work currently? I was under the impression that Docker for Mac already ran containers in an x86 VM. Probably outdated info, but I’m curious when that changed.

Unclear if this question is about Atlas or Google Chrome /s

I had to skim through this a couple times before I realized that I still need to run an MCP server locally. This is basically a proxy between an LLM and the proposed protocol.

It’s a nice proof of concept.

And makes sense that the goal would be for LLM clients to adopt and support the standard natively. Then the proxy won’t be necessary.

There is a standard algorithm - CLDR collation. There are several options available but, generally speaking, it’s a standard.

The specific option for numeric sorting is “kn”.

As far as I can tell, every operating system and many other interfaces tend to use this standard algorithm.

https://www.unicode.org/reports/tr35/tr35-collation.html#CLD...

I think that last line is getting to the point more than most of the discussion I see.

The host of the show doesn’t really decide what he’s going to say. He reads the script the writers deliver (with his approval, I’m sure). He’s the talking head of a production team.

Jimmy Kimmel the person didn’t get taken off the air. Jimmy Kimmel the show did.

That makes it relatively easy to go back on the air, if they simply give new direction (constraints) to the writing team that satisfy the network.

If I were to guess, I doubt Jimmy Kimmel the person cares what the team is or is not allowed to write in the script.

To your second point, the average reader is not a scientist (wrt this topic, at least). Scientists equate proving with knowing. To the average reader, however, there is little correlation between what we know and what we’ve proved.

The scientists don’t know why it happened, because they haven’t proved why it happened.

I’d wager that the average reader knows perfectly well why it happened.

I read it as assuming the deterministic checks and balances are already absent. We have the illusion of determinism but, in practice, audits (and justice) are mostly theatre as it is.

From that perspective, lowering the quality of something that is already non-rigourous might not have any perceivable effect. It’s only a problem if public perception lowers, but that’s a marketing issue that the big-4 already have a handle on.