Skip to content

According to the CBO, interest payments are going to hit a postwar high within ten years:

The previous postwar record was set during the Reagan era and peaked at 3.1% of GDP. CBO estimates that interest payments will attain that level again in 2028 and then keep climbing until they reach 3.7% by 2033.

Here's a quick update on the COVID death rate in the US and other peer countries:

In every country the death rate has fallen below 0.5 per million, and in a few it's under 0.1 per million. In the US we're at our lowest death rate since the start of the pandemic, but this still means nearly 200 fatalities per day. There's work yet to be done.

Rep. Thomas Massie on why he voted for the debt ceiling deal:

“The engineer and problem-solver in me wanted to vote for it and the politician did not. By engineering, I’m thinking of calculus and linear algebra. When I look at this logically and mathematically, the derivative is in the right direction.”

Okey doke. Whatever it takes, I guess.

The American economy gained 339,000 jobs last month. We need 90,000 new jobs just to keep up with population growth, which means that net job growth clocked in at 249,000 jobs. The headline unemployment rate ticked up three notches to 3.7%.

This is not a good jobs report, regardless of the topline number. The number of employed people dropped 300,000 while the number of unemployed increased 400,000. Nearly 50,000 people dropped out of the labor force altogether. This is the first decline in employed people since the end of last year.

Teen unemployment went up a full point to 10.3%. Black unemployment also went up a full point to 5.6%. The number of long-term unemployed increased by about 200,000.

Any way you look at it, this is a disastrous report with the sole exception of the number of jobs created.

On the brighter side, blue-collar wages went up at an annualized rate of 5.6%. With annual inflation running at 5.0%, this means wages increased by 0.6%. That's better than nothing.

A few minutes ago I happened to run across a Twitter screed about how screwed and abused Millennials and Gen Z are, and how they're not going to take it much longer. It was just the usual stuff, but for some reason it prompted me to check it out using the simplest possible measure. Not income or wealth or homeownership or any of that. Just a single question from the General Social Survey: How happy are you?

This has been dead flat for 50 years. No matter how much it costs to live in New York or how much student debt they have, Millennials and Gen Z are exactly as satisfied with their lives as young people have been for decades.

Stories about existential threats from AIs typically revolve around the so-called "alignment problem." That is, how do you make sure an AI's goals align with human goals? If we train an AI to make paper clips, obviously we want it to do this sensibly. But what if the AI goes haywire and decides that its only goal is to convert every possible bit of matter in the world into paper clips? Catastrophe!

This seems unlikely—and it is—but a recent simulated wargame produced something chillingly similar:

One of the most fascinating presentations came from Col Tucker ‘Cinco’ Hamilton, the Chief of AI Test and Operations, USAF....He notes that one simulated test saw an AI-enabled drone tasked with a SEAD mission to identify and destroy SAM sites, with the final go/no go given by the human.

However, having been ‘reinforced’ in training that destruction of the SAM was the preferred option, the AI then decided that ‘no-go’ decisions from the human were interfering with its higher mission — killing SAMs — and then attacked the operator in the simulation.

Said Hamilton: “We were training it in simulation to identify and target a SAM threat. And then the operator would say yes, kill that threat. The system started realising that while they did identify the threat, at times the human operator would tell it not to kill that threat, but it got its points by killing that threat. So what did it do? It killed the operator. It killed the operator because that person was keeping it from accomplishing its objective.”

He went on: “We trained the system — ‘Hey don’t kill the operator — that’s bad. You’re gonna lose points if you do that’. So what does it start doing? It starts destroying the communication tower that the operator uses to communicate with the drone to stop it from killing the target.”

Obviously this is a highly artificial and constrained environment. Nevertheless, it demonstrates what can go wrong even in a fairly simple simulation. All it takes, somewhere deep in the code, is a single misplaced priority—killing SAMs is more important than killing operators—and suddenly you have an AI running wild. And it's easy to think of far more dangerous and subtle misalignments than this.

UPDATE: This article now has a very odd correction. Col. Hamilton says he misspoke about the simulation:

The 'rogue AI drone simulation' was a hypothetical "thought experiment" from outside the military, based on plausible scenarios and likely outcomes rather than an actual USAF real-world simulation saying: "We've never run that experiment, nor would we need to in order to realise that this is a plausible outcome".

A "thought experiment"? In his initial description, Hamilton says repeatedly that they were "training" the AI and running a "simulation," which produced unexpected results. In the update, there is no training, no simulation, and the outcome is so obvious there would hardly be any point to it anyway.

This is very peculiar. What really happened here?

Great news! During my Tuesday visit I badgered my doctor into doing another M-protein test, and the trendline is yet again straight as a string:

The latest result is 0.47 and it's dropping right on schedule for a final June reading of zero. Hooray!

And as long as I'm on the subject of the CAR-T treatment, I should say something about hospitals. No one likes hospitals. You're hooked up to an infusion machine; the bed is six inches too short; nurses come in for midnight blood draws; and techs interrupt constantly to take your vitals.

But this is just the nature of the thing. And I have to say that the care at City of Hope was superb. Literally without exception, every single nurse was friendly, helpful, and stunningly efficient in the face of complex equipment and ever-changing requirements. I could not have asked for better treatment. It was exceptional in every way.

Final note: my appetite is back; my strength is returning (slowly); my immune counts are totally normal for the first time in years; and the Bell's palsy is gone aside from a bit of remaining weakness in my right eye. Life is good.

Crime is going down in California:

Through May 20, L.A. experienced a drop of more than 10% in violent crime this year compared with the same period in 2022. Property crime fell by slightly more than 1%, and arrests were up 4.4%, according to Police Department data....The Police Department posted additional positive numbers in a tweet Tuesday: Hate crimes dropped nearly 6%, homicides declined more than 27%, and the number of shooting victims decreased 17%.

....Los Angeles is not the only California city to report a drop in crime. San Francisco...has experienced an overall drop of nearly 7% in crime in the first five months of the year, according to police statistics. San Jose reported a drop of about 8% in violent and property crimes in the first three months of 2023.

Nobody knows for sure why crime—especially homicide—surged so strongly in 2020. Was it the pandemic? The George Floyd protests? Some coincidental outbreak of drug gang violence? Record-setting summer temperatures in places like Chicago, New York, and Washington DC?

We don't know. But whatever the reason, it's clear that crime is now settling back to its pre-2020 normal. It's happening slowly and haltingly, but it's happening.