rhplus

Their contention seems mainly focused on the heavy resource use of AI tooling.

They run a service: they could impose limits on the number of submissions, builds, branches, scans, etc, and give additional usage credits to the large community projects they want to encourage.

What happens to all the precious metals that go up with these data centers? At least with a land based data center we have a vague hope of recycling the raw materials at some point in the future.

Pleasant flashbacks to the Windows 95 classic Serene Screen “Marine Aquarium”

https://youtu.be/PprsZgA-kxs?si=waSpdI4J323fNNGh

https://www.pcgamer.com/gaming-industry/a-former-air-force-p...

I think this article highlights more the importance of understanding system limits in the 1990s versus today. No-one would care to much today if the code review for this feature had “files.count()” or whatever in it, but in the mid 90s that would have been a huge performance red flag because a user would literally hear their hard drive clicking away and see the blinkenlights.

Kind of out there, but if you were running a network of agents that are autonomously looking for exploits, then perhaps those agents have decided that a good source of fresh ideas might be to periodically receive Schneier’s newsletter in an inbox they control.

> not always your first exposure

Great point. How many professional sports players credit their elementary aged PE classes to their success in sport (as opposed to say crediting a high school coach)?

Progressive, unbridled access and exposure is key.

> unenumerable amount of knowledge, tips and experience silently disappears

Not to be negative about experts here, but of all industries, isn’t airline maintenance one of the most documented and standard procedure focused of them all? I would hope that someone maintaining a 777 doesn’t have to remember certain tricks or tips outside of what the manuals say, and when mistakes are found, they’d be noted, investigated and republished to all maintenance crews worldwide.

I sense this is not winnable for all Wordle dictionary words. Is that provably the case?

Don’t forget operations and maintenance. You can have a box of vibe coded custom tools and services, but they’re all going to slowly rot as the underlying dependencies - network APIs, platform APIs, auth handshakes, reverse engineered file & DB schemas - change their behaviors and availability.

Sure, an LLM can fix it, but that assumes your original spec sufficiently described the user requirements and wasn’t just a transcript of the original session trying to figure out how how to interface with all those dependencies in the first place.

> Touching face, chest, groin, etc is forbidden.

It’s really hard to believe that an officer who’s wrestling someone hand-to-hand, possibly on the ground or against a wall, is going to restrain themselves from grabbing face or chest in the heat of the moment.

The first half of the article seemed fine, but then it appeared padded with AI filler.

They even predicted that smartphone users would inevitably ruin the (movie) theater experience (skip to 3:55 and 4:57)

> Then LLMs killed stackoverflow.

I think the common consensus is that stackoverflow killed stackoverflow, quite a few years before LLMs became entrenched.

Especially the pill counting example. The best model was shown at 81.1% accuracy, which is a terrible rate for pharmacy scenarios. It seems like implementors would be better off instructing the models to use deterministic tools (like OpenCV) until the models are at 99.99% accuracy (or whatever an acceptable error rate is for pharmacy techs).

I’m fairly certain the Tesco and Sainsburys near me were still using vacuum tubes to send money and receipts up to their back-offices well into the 1990s.

I don’t think the emphasis here is on real time multiplayer, but instead on async multiplayer. The way it’s described is about an evolving code base over time. Everyone sees the conversations, comments and decisions that lead to the code in screen. Individuals directly interact with the code via agents. Other devs see all those deltas and can add to the discussion.

It looks like you’re describing something like an MCP server and a client model.

If you’re in the C# ecosystem you could consider converting your APIs to MCP format tools using the MCP SDK.

https://devblogs.microsoft.com/dotnet/build-a-model-context-...

And then leveraging Microsoft Agent Framework for the client and orchestration side of things:

https://learn.microsoft.com/en-us/agent-framework/

A parked domain would potentially make more money hosting AI generated “content” to improve its SEO and passive ad revenue. That’s why invested interests have pushed for this side-channel “For sale” sign.

For iPhone users, you can set up a Shortcut that achieves similar results because many snatchers will instinctively try to turn off cellular:

+ When airplane mode is enabled + Lock the screen and require password

Locking the phone is the most important thing. Losing the physical device is a bummer but it can be replaced and restored. Allowing access to any apps or message history is the worst case scenario.

Unfortunately, they broadcast on DAB so there’s a strong chance that two car radios next to each other are buffering differently and out of sync.

Your ratios are way off. Price ratio is 1:1500 for A to B, while voting power ration is 1:10000. Each A share is 6.67x more powerful than 1500 x B shares.

Some media companies had them before the 1980s. New York Times issued dual class in 1969 so that the owning family maintained editorial control.

Berkshire Hathaway is possibly the most famous from the 80s/90s. The class A shares are significantly more expensive and proportionately even more powerful than the class B shares. The lower price version was important back when physical exchanges didn’t support fractional shares as they do today.

Disappointed that xn--sei.com or <insert fleuron here>.com is apparently registered but not redirecting to ornately decorated texts.

Images, music, video, and text would all be under copyright, while characters and logos may be registered trademarks.

XKCD: 10,000 people learn something “everyone knows” every day:

https://xkcd.com/1053/

Someone has an ambiguous bet predicting when RFC 10000 will be published, but the numbers went straight from 9998 to 10008. No-one wins!

https://manifold.markets/CollectedOverSpread/when-will-rfc-1...

BS check: $8T is $1000 per person on this planet. A healthy P/E ratio of 25 would translate to earning $40/year in profit from every person on the planet. SpaceX/Starlink obviously doesn’t just walk in and get everyone as a customer though. They have roughly 10 million customers right now. Let’s be generous and say they have 20 million. That $8T works out at $400K per customer valuation which at a 50 P/E would mean $8K/year/customer profit per customer or $666/month/customer profit. Those are generous numbers. Scaling back to 10 million customers and 25 P/E would require $2666/month/customer in profit to get to $8T valuation. For Internet service?

7/11 Japan really benefits from urban density, which in turn makes the distribution of fresher food and smaller footprint stores much more of a factor.

The distribution network even shows up in maps. There will be clusters of 7/11 in Japanese cities which is more efficient than spreading them equally.

https://conbini.kikkia.dev/

Billable hour rates would need to increase by 25%.

Most businesses don’t grow just by churning out more units of software. At some point, it doesn’t matter how quickly you can churn out features if you’re not solving customer problems and convincing customers that they should pay for those solutions.

Once software becomes cheap, the bottleneck to growth shift to product design, infrastructure/manufacturing, sales and support.