Agreed - Rozado’s been doing very interesting work.
I could have been clearer in my choice of words - haven’t yet seen/heard of instances of “LLM manipulation” as in the LLM attempting to manipulate its audience. That could be a scenario in which, for instance, the LLM gaslights the user.
South Korea has the fourth-longest working hours among countries in the Organization for Economic Cooperation and Development (OECD), with employees working an average of 1,915 hours per year — compared to the OECD average of 1,716 hours.
which makes one wonder how other countries rank in average hours worked. Here are the data, as published by the OECD.
Measuring average hours worked per year is more representative than length of work week because it takes into account factors such as holidays, sick leave, etc. Here is the complete definition.
Average annual hours worked is defined as the total number of hours actually worked per year divided by the average number of people in employment per year. Actual hours worked include regular work hours of full-time, part-time and part-year workers, paid and unpaid overtime, hours worked in additional jobs, and exclude time not worked because of public holidays, annual paid leave, own illness, injury and temporary disability, maternity leave, parental leave, schooling or training, slack work for technical or economic reasons, strike or labour dispute, bad weather, compensation leave and other reasons. The data cover employees and self-employed workers. This indicator is measured in terms of hours per worker per year.
One of the biggest new initiatives in the budget is the ISS deorbit tug, which would be used to perform the final lowering of the station’s orbit to ensure it reenters over the South Pacific. NASA first indicated its plans for the tug in a request for information last August, but offered few specifics about the vehicle in the budget request.
The $180 million NASA is requesting for the tug “gives us a healthy start” for the project, said Kathy Lueders, NASA associate administrator for space operations, in a media teleconference about the budget.
While budget documents did not include a spending profile for the project, Lueders said the agency came up with a cost estimate “a little bit short of about $1 billion.” The exact amount, she said, will depend on what proposals the agency receives from industry from an upcoming request for proposals (RFP).
“Our goal is to go out with an RFP,” she said. “We’re hoping to get a better price than that.”
So, around US$ 150 billion to build it, around US$ 3 billion a year (NASA expenditures only) to operate it, and another billion to burn it up in the atmosphere and drop the chunks that survive reentry into the South Pacific abyss.
In my limited experience visiting some factories and design centers in China, they kick you out of the office after 10pm, one needs a permit to come to the office on Saturday and it is nearly impossible to obtain a permit to come to work on Sunday.
It would take a lot more than a billion. The mass of the ISS is around 420 tonnes (not counting any modules which may be added before the disposal is performed). The delta-v to get from low Earth orbit to an orbit at the lunar distance (which is what you need to get to L4 or L5) is around 3.26 km/sec. Plugging this into the rocket equation, and assuming you’re using LOX/liquid hydrogen fuel, and neglecting the mass of tankage, engines, and structural components, the initial mass would be 881 tonnes, or a booster of 461 tonnes. If we assume something like the Saturn V’s S-IVB third stage is being used as the booster, it would take more than six Saturn V launches to lift the stages into orbit, and that’s ignoring how they would be coupled together as a propulsion unit for the station.
It would be much less expensive to boost the ISS into a higher Earth “graveyard” orbit where it would be safe from atmospheric decay. If it were boosted into a 900 km circular orbit, its orbital lifetime would be on the order of 1000 years, which should be plenty of time to decide what to do with it next. If Starship becomes operational with orbital refuelling (which is required to perform the lunar demonstration mission which NASA has scheduled for 2024), it should be able to perform that boost operation in a modest number of flights.