in IT. <rant>My current salary is in the top 10% vs the US CS average. My 1st job paid $25K. Since then, the longest stretch of unemployment I've had was 3 weeks. I've never needed to look for a job, even after quitting an idiot boss during the 2008/9 recession. I've had 3 messages on my voicemail this week from recruiters asking if I was available. I've saved for 30 years and I can retire anytime I get bored with work.

My 25 year old daughter just got a $50K/yr IT position with our local credit union here in Ohio after working there a year as a teller. Her bonifides were a series of classes from Coursera resulting in full stack development certs from Meta, IBM, and Google. Ok, she did have her Dad as a resource when she got really stuck, but she did the work herself. She'll finish her associates degree from the local community college, get into a 4 year degree program, and be on her way. I'm not at all worried about her future.

I graduated from college in 1983. The US unemployment rate hit 10.8% that year with inflation running at 3.5% (inflation was 13.9%/yr halfway through my undergrad degree). I don't feel all that much sympathy for new grads who don't know how to network, cold call companies and then complain about a job market when the national unemployment rate is a mere 4.2%. That is pretty darn close to full employment.

I swear, kids these days...</rant>

I grew up in Oak Ridge, TN. I knew many people who worked on the bomb, mostly on the uranium isotope separation problem. I'm 2 degrees of separation from Oppenheimer, Lawrence, Teller, etc.

My Father was stationed on Okinawa during the Korean War. He shot photos of some of the caves where the bones of thousands of Japanese soldiers and civilians who died during Okinawa were stored. I've read that the average lifespan of a 2nd Lt of Marines who hit the beaches of Okinawa was measured in minutes. A friend whose father served in the Marines and who arrived in Okinawa after the battle in preparation for the invasion of Japan, claims that the military ordered a quarter million Purple Heart medals they expected to award in the battle for Japan. We are, my friend tells me, still giving out those medals today to soldiers wounded in Iraq and Afghanistan.

There is now an International Friendship Bell in Oak Ridge, located across the Turnpike from where I went to high school. The bell symbolizes peace and unity and was rung on the 80th anniversary of the Hiroshima bomb. But the bell was somewhat controversial in Oak Ridge since many regarded it as an apology for the role Oak Ridge played in the bombing of Hiroshima. The scientists and engineers I knew who worked on the Manhattan Project had few moral problems with the use of the 1st atomic bomb on Hiroshima.

There is a philosophical question in there somewhere. If the simulated lifeforms survive from reboot to reboot, could they use observed differences in the physics from reboot to reboot to detect that they were, in fact, simulated life forms? To prevent that, wouldn't the current simulation software version require that experiments run in an earlier version of the simulation under the old physics continue to produce the same old contradictory answers?

I'd guess that you have never served in the military and thus have no idea how ridiculous it is to believe that a handful of 2nd Amendment kooks armed with AR-15s can effectively stand up against a modern army.

It is unclear what the ROI is on all of these AI datacenters being built. This seems highly reminiscent of the fiber buildout during the internet bubble. Meanwhile, US copper prices jumped 13% today after Trump's announcement of a 50% import tariff. I'd imagine that will up the investment cost of building still more datacenters. At some point, some catalyst like copper prices will cause all of this to implode and Nvidia, currently priced to perfection, is going to take a hit.

The bigger they are, the harder they fall.

Somewhere I've read where that has been tried. The result was a language that was incomprehensible to humans. Bad enough that LLMs and other neural nets are essentially black boxes where the inner workings are difficult to decipher. I'm not sure I want black box LLMs communicating amongst themselves using a black box protocol. Debugging that would be a nightmare.

I thought UN*X was built on the notion that a byte stream was a byte stream was a byte stream. Plain text is just one kind of byte stream among many.

You need to upgrade the esr once a month or more anyway just to fix the latest batch of security issues.

John McEnroe judge.

No. Retires are *NOT* part of the labor force. They are counted as part of the denominator (the population) when calculating the labor force participation rate. The labor force itself is the numerator which consists of persons over the age of 16 who are either actively working or looking for work. Retirees are neither. See the Bureau of Labor Statistics for the precise definitions. Econ, like all science, requires precise definitions of concepts in order to measure anything or test hypotheses.

I'm not the poster who claimed the 37% unemployment figure.

A grad student with a research or teaching assistantship has a job. My point is that such job opportunities, which have helped many people get through recessions, are no longer available because of the cuts that Trump/DOGE has made. Many universities have had to tell accepted applicants that that their offers have been rescinded. See, for example.

I would guess that you have never been to grad school.

Retired folks are not considered unemployed. They are simply excluded from the definition of the labor force. The labor force participation rate is equal to (the number employed + the number actively looking for work) / population. To be considered as unemployed, a person must be actively looking for work. Similarly, discouraged workers who have quit looking for work are also excluded from the labor force definition.

Not my point. DOGE has had significant effects on the availability of the research dollars that support grad students. That prevents recent college grads from going to grad school the way I did after my undergrad degree.

The Saudi oil embargo and the advent of OPEC had the biggest effect on oil prices in the 70s.

It was Arthur Burns, Volcker's predecessor, who failed to raise rates on time to quell the inflation of the 1970s. Volcker raised rates to 20% to stop inflation. It did, but caused a nasty recession just as I was graduating.

Keynes would disagree. The policies of FDR, LBJ, Nixon, Carter, Reagan, GW Bush, Obama, and Biden have all had significant impacts on the economy. OTOH, for whatever Trump's policies have done to the stock market, they have barely registered on GDP growth, employment, and inflation so far. We won't know what real effects his policies will have for several more months.

BTW, my undergrad and grad degrees are in Economics. I was in the US Army when the Iranians seized the US Embassy. I recently realized that I've lived through 25% of US history since 1776.

The unemployment rate was 10.8% when I graduated in 1983. Fortunately, grad school offered a research fellowship which covered tuition, paid the rent, kept food on the table. and allowed me to ride out the Reagan/Volcker recession. But DOGE is killing that option today.

Statisticians old enough to have used cards will remind you that you do *NOT* always want to resort your cards after you have dropped them. "Accidentally" dropping your data deck down the stairwell is a time-honored solution to autocorrelated errors in time series data. Unfortunately, though, since the bigger numbers have more ink, they will travel farther and so, if you pick them up in order starting from the top of the stairs, you are likely to find the data may now have heteroscedastic errors.