e_y_

"Engineers are more likely to be terrorists" is different from "terrorists are more likely to be engineers".

You can imagine that engineering is a useful skill for terrorism and thus terrorist organizations might spend extra effort trying to recruit engineers. They may also have a higher survival rate working on behind the scenes tasks rather than firefights and suicide missions, which could cause a survivorship bias in data collection.

(It's also interesting how many foreign leaders and dictators have engineering or science degrees, and/or went to US universities prior to becoming leading figures in their home countries.)

CNN quoted someone from Google saying it was a fake (old account that just changed its handle today).

IMHO there ought to be restrictions and transparency on channel name changes because it gets abused all the time.

The guy's face has been plastered over the news for several days and there's a $60,000 reward. Getting a tip from a fast food worker is very plausible.

More plausible in my opinion than the FBI having some kind of agreement with McDonald's to access their store surveillance network in real time.

The robot arm isn't running off a battery and most of them are large enough that board size isn't a constraint, so why add the complexity of custom silicon if it isn't needed? Using off-the-shelf FPGAs also makes repairs easier.

I think adding labels to the product is only useful to a certain extent. There's a few things that stand out, like if a product has 70% DV sodium that's usually a bad sign, or "cheese product" tells you that it's not real cheese (for better or worse). But stopping to look at the ingredient list of every product you buy is time consuming.

This is where online food databases might work better. Scan the UPC/EAN with an app and the app can tell you which ingredients are generally safe, which ones to avoid. The recommendations could be personalized to your dietary needs or allow ratings from different organizations that might have varying criteria for what they consider OK vs bad.

Such apps exist but usually requires someone (volunteer crowdsourcing or paid) to input products into the system so it may not be complete or fully up to date. And as you suggested, supplier information and more detail on the processing isn't available. Having an international, public database of that information (with an API for app makers) could help make nutrition label apps -- or maybe even built into Google Lens -- much more accessible to the general public.

That's probably where there's a big advantage to being a company like Nvidia, which has both the proprietary chip design knowledge/data and the resources/money and AI/LLM expertise to work on something specialized like this.

1. Before the pandemic, farebox recovery was 73%, one of the highest in the country. Electrified, it probably would have gone even higher since at least some people avoided the train because it was too crowded at rush hour. Unfortunately it could take over a decade to return to that point if ever, but it shows that Caltrain can be fairly efficient given enough ridership.

2. There have been ongoing efforts to improve equity, including Go passes for students and low-income workers. Clipper START offers 50% off for low-income riders, youths can ride for $2/day now. Side note: subsidies and programs like this decrease farebox recovery ratios, which is why transit covering its own costs is not necessarily the end goal. In fact, the decline of commuting for tech/knowledge workers (many who now work hybrid or fully remote) has meant that less-privileged workers have started making up a larger share of Caltrain ridership.

3. The new electric trains are a lot quieter. There's ongoing, long-term projects to grade separate the remaining at-grade crossings (of the 113 crossings, 41 still need to be separated).

4. Arguably, the stations were directly responsible for the land being so valuable. Many peninsula downtowns and sometimes entire cities were built around the train stations, 100-150 years ago. Pre-pandemic, some of the most expensive apartments were the ones near the train station. Sure most of them also had the benefit of being close to downtown restaurants, but there were also big developments around Lawrence station which has basically zero local services aside from the Costco and the train station.

Of course this was mostly in the past, but downtown areas continue to be congested and the only way we can keep building high-density housing is to have good public transit. In the long run, the train provide immense benefits to the cities that host them and removing it would be shortsighted.

According to rumor the AEM-7s were fried during testing due to a misconfiguration.

In fact it was looked into by Caltrain's predecessor, the Southern Pacific railroad

https://cdnc.ucr.edu/?a=d&d=SDI19210907.2.17

Stockton Independent, September 1921

> Electrification of Lines to Be Looked Into by S.P.

> SAN FRANCISCO, Sept. 6. — William Sproule, president of the Southern Pacific company, has received authorization from the board of directors of the railway to investigate the feasibility of electrifying the Southern Pacific lines between San Francisco and Peninsula points, It was announced here today. Better and more frequent service would result from electrification of the peninsula lines, according to parties who petitioned the Southern Pacific to take this action.

(Unfortunately, it was dismissed a couple weeks later: https://cdnc.ucr.edu/?a=d&d=SDI19210917.2.13)

Possibly there's some novelty factor to riding the new trains. My 70-something year old aunt, visiting from another continent, wanted to ride the new trains (and they weren't even fully converted yet, it was day 2 of the soft launch and only a couple trains had been switched over so I had to check the train sightings on Discord). Obviously they're not all that unique outside the US but it's exciting to see them here.

But weekend service is now twice as frequent (now every 30 minutes) which is a game changer, so it's not a surprise that those numbers in particular have shot up since they were pretty weak to begin with.

As I understand it, they don't own every stock. Right now VTI has about 3700 stocks out of the ~4300 on US publicly traded markets (excluding OTC). But it's more representative than the S&P 500 and similar.

I would agree that bytecode is different from machine code, but "efficient execution by a software interpreter" seems like a bad delineation considering that some virtual machines treat bytecode as simply an intermediate representation for JIT. WASM bytecode is very similar to machine code and meant to be translated to machine code rather than interpreted.

I would define bytecode as:

- Closer to machine operations than something like an AST, which is where I think the "efficient execution" part of the wikipedia definition came from

- Not directly executable on any physical CPU, or at least not typically. There are some CPUs that can execute Java/Python bytecode (sort of) but they're in the minority.

While not strictly required, bytecode tends to have opcodes meant to be used with a virtual machine, such as dealing with objects instead of linear memory.

This is all to say, the proper terminology for RISC-V is an ISA, but the distinction is more about how it's being used than how it's designed.

Not quite haunted but I've had people report that my website hosted on a .quest domain is blocked on their work computer. My best guess is that their filter thinks it's gaming related (it's not) or maybe they just block all "weird" domains.

My uninformed guess is that aside from the obvious glaring user base size problem, the API is too locked down to build unique, interesting apps and too different/incompatible to port existing ones like VR games.

I learned from Scott Manley's latest video that buoy is pronounced "boy" in British English (boo-ee in American)

The turnaround time for Shuttle was 2-3 months, while building a brand new rocket takes like 1-2 years. Although for the cost of Shuttle, we could have built a whole bunch of expendable rockets, pipelined with a regular launch cadence, and probably also gotten some cost savings through economies of scale.

(Not an engineer either) My understanding is that it's been done before on smaller scales but having a giant piston in the tank requires a good seal, railings to keep it straight, and overall way too much mass and rigidity. Consider that the tank walls are only a few millimeters thick.

Instead, it's more common to use gasses injected at the top of the tank to push the liquid to the bottom. Falcon 9 uses helium. Starship uses https://en.wikipedia.org/wiki/Autogenous_pressurization as well as small header tanks for the landing propellants.

A system (whole) that can survive the failure of (some of the) individual parts. Up to a limit.

We're not doing that while driving, though.

If we're talking purely about going off memory, there's no reason why machines couldn't build up a similar catalog (which could be used by every self driving AI once learned). And human ability to judge distances varies significantly between drivers.

That very much depends on the game. Multiplayer games are at the highest risk of becoming unplayable. Most single player game should in theory survive the death of the developer/publisher as long as Steam is around.

Some single player games require a central server for no particularly good reason, or for some functionality that could be disabled without much loss to the game's playability. I think these are the ones that should be called out, and required to either have an end-of-life plan (a flag ready to go that disables online functionality) or a big warning "whelp this game will stop working eventually buy at your own risk".

Citation? Humans are not constantly moving their heads to the degree that chickens do, and I find it doubtful that the micro movements from our head (which our eyes have to adjust for with the vestibulo-ocular reflex so things aren't blurry, similar to image stabilization in cameras) are large enough to infer depth.

This is not a No True Scotsman fallacy ... and if you argue otherwise, you're falling for the No True Scotsman fallacy.

Just kidding. But more seriously, the original claim was deeply flawed so it makes sense to challenge the criteria for the population study.

At my company, DB migrations on startup was a flag that was enabled for local development and disabled for production deploys. Some teams had it enabled for staging/pre-production deploys, and a few teams had it turned on for production deploys (although those teams only had infrequent, minor changes like adding a new column).

Personally I found the idea of having multiple instances running the same schema update job at the same time (even if locks would keep it from running in practice) to be concerning so I always had it disabled for deploys.

I thought Uber just let drivers ("independent contractors") deal with the vehicle maintenance and cleaning up when someone pukes in the back. Do they even have their own depots?

LLM's general knowledge is trained on data that's months old (or longer). Alex Morgan only announced her retirement a few days ago, so in order for the AI to know about it, someone would need to pick out recent, relevant articles about her and feed it to the generator.

And even then, it's unclear that the AI would be smart enough to identify "retirement" as something that should be called out in the new article without specific prompting.

The amount of money and budget that a company has can vary wildly. Just because your company uses SSO doesn't mean you're a Fortune 500.

Of course this mostly means that if your required feature set falls into the "Contact us for pricing" tier then you're going to have to talk to a sales rep (probably several, from different vendors so you have a competitive quote) instead of getting any kind transparent pricing.

I think they even robocall stores with something similar to Assistant to check if they're open on holidays, which is definitely something that would be difficult for a smaller or community-run database to accomplish at scale

We should be doing both, for different reasons. Ban data brokers because they allow anyone with a credit card to stalk people, more or less legally. Fix the SSN identity system because even if you ban data broker businesses, dark web brokers don't abide by the laws anyways.

Maybe the old days when you needed multiple maps folded up on paper instead of one zoomable super map containing the entire world.

Also "Google Map" sounds like a map of Google.

In the press conference, they said the collision risk can be avoided by undocking the Starliner and then letting it float away to a safe distance before starting up the thrusters.