They're not saving money. They're retasking office workers who make $100+ per hour to do work they usually pay a lot less for.

OTOH, if it keeps customers from having bad experiences because the system is overwhelmed, it may be a good use of those expensive workers.

Nah. These are salaried workers being asked to do something during their normal work hours. It's basically not possible to avoid paying them.

Weekdays 10 am to 6 pm. That's normal work hours. Unless they're being forced to use vacation time, they're being paid.

More likely 100-200/hour, more if they're software engineers or similar highly-paid office workers. $30/hour is only $60k/year. There's no way Amazon white-collar workers in NYC are making that little.

Thanks. It doesn't appear in my Apps list, so I probably took it out via the ADB interaction. However, I don't like seeing it in the list of possible "Digital Assistants" if it's been uninstalled.

Oblig. Dr. Ian Malcolm: "Yeah, but your scientists were so preoccupied with whether or not they could, they didn't stop to think if they should."

That's good to hear!

Now, how do I rip out Gemini, or ensure it's never installed?

What, in your argument, is the difference between LLM copy-edited text, and for-hire human copy-edited text. The editorial services I have seen *sometimes* try to find editors that are kinda-sorta near the correct field of expertise, but there's no guarantee you'll get someone who even has a passing level of familiarity with your field, and for some services, all they have is a degree in English.

So, again, what's the difference between linguistic polishing by machine and linguistic polishing by semi-qualified human?

Following up on that idea, there are various copy-editing services that many non-native English speakers use, and are encouraged to use, to help improve their writing. The main difference from the perspective of forensic detection with AI-copy-edited text is that there are a very small number of such styles compared to the likely thousands of copy-editors' individual styles, making automated copy-editing easier to detect. I'll bet dollars to donuts that if you trained an LLM on the output of a single human copy-editor, you'd be able to identify all papers that used their services.

We use AI to help with paper writing in my lab, mostly because there are only two native English speakers, and it relieves me, the lab head (and one of the two native speakers), of having to do extensive copy-editing in order to make stilted English more readable. I still read every word that gets published from the lab, but using AI for copy-editing is no different from using a human-based writing service to fix poor language. It's just cheaper and orders of magnitude faster.

So, for us, the response would be a big, "so what?" to this report.

But, if people are starting to use AI to write entire papers, that's a different story. My experience is that current models hallucinate ideas and, especially, references, at far, far to high a rate to be seriously useful as anything other than a tool that requires full, manual verification. I half-jokingly say that if a paper is hallucinated, that means the AI was unable to find the right citation, and it represents a gap in the field's knowledge that we could address. The amazing thing about the hallucinations is how convincingly real they sound: the right authors, the right titles, the right journals. These are publications that *should* exist, but don't, at least in my experience.

As a most recent example, when writing a grant application, I tried to find citations using an LLM for an idea that is widely-held in the field. Everyone knows it to be true. It's obvious that it should be true. And, yet, there have been no publications as of yet that have actually discussed the idea, so the LLM dutifully hallucinated a citation with exactly the author list you would expect to have studied the question, a title that hits the nail on the head, and a journal exactly where you might expect the paper to appear. I've told my staff that we need to get that paper written and submitted, immediately, to fill that obvious gap, before someone else does. It will likely be cited widely.

The 677K scrapped were not nearly enough to affect national used car prices over time because that was a drop in the proverbial bucket among tens of millions of new and used vehicle sales and scrappage. CFC did not target ALL used vehicles but select models.

Meanwhile the US total car fleet shrank by FOUR MILLION in 2009 (reflecting the 2008 recession far more than CFC

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.thedrive.com%2Fnews%2F...

Meanwhile the 2009 total scrapped was over FOURTEEN MILLION.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fgrist.org%2Farticle%2Fu-s-...

"In 2009, the 14 million cars scrapped exceeded the 10 million new cars sold, shrinking the U.S. fleet by 4 million, or nearly 2 percent in one year. While this is widely associated with the recession, it is in fact caused by several converging forces.

Future U.S. fleet size will be determined by the relationship between two trends: new car sales and cars scrapped. Cars scrapped exceeded new car sales in 2009 for the first time since World War II, shrinking the U.S. vehicle fleet from the all-time high of 250 million to 246 million. It now appears that this new trend of scrappage exceeding sales could continue through at least 2020.

Among the trends that are keeping sales well below the annual figure of 15â"17 million that prevailed from 1994 through 2007 are market saturation, ongoing urbanization, economic uncertainty, oil insecurity, rising gasoline prices, frustration with traffic congestion, mounting concerns about climate change, and a declining interest in cars among young people."
---

Being a Boomer I should stereotypically believe unsubstantiated drivel propagated by industry-illiterates but it happens I've been wrenching including scrapping cars and harvesting parts from salvage since the 1970s including the CFC era. Facts are fun. I recommend inquisitive learning instead of AssUming correlation is causation.

I read the rules and checked the numbers because we went to auctions where CFC cars were bought by authorized yards where self and boss went to harvest delicious parts (I got all I could pull free if we had room to load them which was nice) some of which came from vehicles like CFC Jeep SUVs (no loss since their transmissions put many otherwise sound ones into salvage and transmissions to repair those on the road were legal to harvest from CFC vehicles). After skinning them out the mandated hull and engine got scrapped with any unwanted leftovers, but the harvested parts kept many others on the road during that recession era.

The number CFC crushed was not even close to enough to change used car prices and (this may be terribly difficult to understand because OMG math is hard) the impact of the tiny few recycled IN 2009 is not what drove used car prices up from then until 2025. (Vehicle price) inflation did because used car prices lag new car prices while driving them upwards. As new vehicle prices rise so do used, there being no reason to sell things for less than market value.

US auto production in 2009 alone was ~5.71 million units (per DOE). Total sales were around ten million for that recession year.

CFC didn't target anything of great cultural value (and though this may be painful for people who REALLY want a special vehicle but not quite enough to do more than blame circumstance for not un-assing the couch and getting one they were too slack to track down as I do) the few pretty toys were exceptions not rules hence irrelevant.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Ffred.stlouisfed.org%2Fse... gives an idea of total vehicle sales over time.

Few understand 2009 also saw roughly 14 MILLION vehicles scrapped. People offed unwanted tradeable and untradeable (a few hundred for scrap cars moves a lot of iron) vehicles after the recession. That's where the "used" cars went in volume enough to briefly ripple the market then inflation did the rest until present day (when nearly every CFC vehicle would have been scrapped anyway).

Most vehicles taken to salvage could be repaired (burn jobs and utterly smashed excepted) but not at US labor rates so auction buyers send parts and complete vehicles offshore where people still know how to work on things, notably sheet metal/paint/interiors (sheet metal and interiors are easy swaps from other donors).

Used cars in the US supply export markets too and the auction system is highly efficient at moving cars/trucks/SUV/construction equipment/big rigs/etc from backwaters to seaports. That's far from new either since new vehicles are a luxury purchase.

2009 exports exceeded CFC, but I don't see anyone boohooing about that though UNlike CFC, exports send more valuable vehicles for the most part since the markup for the reseller is far more than some econobox they can get used from the EU then truck instead of ship by maritime car carrier.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.usitc.gov%2Fpublicat...

"Used Vehicles Are an Important Component of U.S. Passenger-Vehicle Exports

U.S. exports of used passenger vehicles averaged 762,000 per year from 2009 to 2013, with
many destined for lower-income markets. The large foreign market for these exports appears to
have propped up U.S. passenger vehicle sales and may have lessened the price premium for new
vehicles to U.S. consumers. This suggests that the current level of U.S. new-vehicle purchases
can be at least partly attributed to foreign demand for U.S. used vehicles. Estimated data are
used because precise data on some U.S. exports of used vehicles are not available.1
Used vehicles made up 34 percent of U.S. passenger vehicle exports by volume during 2009â"13
(21 percent by value). Major markets for U.S. used vehicle exports differ from markets for new
vehicle exports, particularly in terms of volume. Used passenger vehicles are often exported to
developing countries because they offer luxury options at lower prices, are not available new in
those countries, or match or exceed the quality of locally available new vehicles.
Estimated U.S. exports of used passenger vehicles rose 26 percent from 2009 to 2013, from
656,000 units to nearly 826,000 (figure 1). At the same time, U.S. exports of new vehicles
doubled to nearly 1.9 million in 2013, causing the share of used vehicles among all U.S. exports
of passenger vehicles to fall from 41 percent in 2009 to 30 percent in 2013. The growing global
economy drove a jump in demand for more expensive new vehicles as well as used vehicles.

This just raises the question of what it means to "know". The LLMs clearly have a large and fairly comprehensive model of the world, the things in it and the relationships between them. If they didn't, they couldn't produce output that makes sense in the context of the models we have of the world, the things in it and the relationships between them.

That has nothing to do with the comment you replied to. I was talking about Trump's cluelessness what is needed to retain the dollar's status as the world's reserve currency which is at best weakly related to its valuation relative to other currencies.

Only if hedge fund managers don't understand how the Fed chairmanship works. Trump can't replace Powell until February 2026, when Powell's term expires. Not unless Trump can make the case that he needs to be removed for cause, which would require evidence of misconduct, not just policy disagreement.

I agree entirely. I teach an intro to programming course at one of the well-known universities. It is a lab course with 2 hours of teaching contact time per week, 2 hours of reading time per week, and 8 hours of expected programming time per week. The students learn by doing.

Aw, c'mon. Basic is a horrible, horrible language that teaches poor habits and has almost no translation to serious programming.

If you want to have kids learn something easy that at least teaches good organization and thought processes, then teach them Scheme.