She also tried to kill another prisoner several years after her initial crime. That’s reason to believe she’s more murderous than the average murderer and would present a danger to the public if released.

Life in prison or a mental health facility seems like the only reasonable option to me unless she was able to present compelling evidence that she is no longer dangerous.

Not to mention, it’s a significantly less effective tactic in a ranked choice system.

One of the goals of imprisonment is to punish. I’m willing to agree that she has been punished enough.

Another goal is to protect society from people who have demonstrated an ability and willingness to harm others. I’d be entirely fine with moving Pike to a more comfortable prison or mental institution.

It’s also a normal response for the authorities to try to keep the murderer in prison, even if that means harsher conditions.

I have issues with the prison system in the USA. I think far too many people are imprisoned, and the system fails to rehabilitate people it could and should. That said, I do not want Christa Pike to escape, or be released into society.

She beat a girl to death and kept a chunk of her skull as a trophy, then later tried to strangle another prisoner. She showed herself to be an extreme danger to others and should be confined for the rest of her life, or until there is clear and convincing evidence that she is no longer dangerous.

28 years of solitary confinement sounds needlessly cruel even for someone convicted of a terrible crime, however, it was arguably justified by her actions in prison. She tried to kill another inmate, for which she was convicted, and later, she tried to escape.

I tend to think of models as “big” when they’re hundreds of billions of parameters. GLM 5.x is “big”. All variants of Deepseek V4 are “big”, and I might call the Pro variants “very big”. Kimi K3 is “very big”.

I do not own the hardware to run those locally.

This matches my experience when I’ve used midsize models connected to read-only tools.

It does not match my experience using a big model in an environment where it can generate and execute code when given instructions to base its answer on that. That pattern almost always results in a factual answer, and often, but not always a good one. The downside is you need to set up such an environment or pay a provider for one, and it’s slow. If you have something else to do while you wait, it may improve your efficiency. If you don’t, you might as well do the experiment yourself.

The opinion I’m starting to settle on is that expecting people to read LLM output is rude.

That means at the point something is released for public consumption, the documentation should be human-written. At the point an open source project solicits collaborators, the code should at least be human-reviewed and probably human-refactored. I’ve read enough LLM-generated code to know it isn’t very good at picking abstractions.

There are, of course business school types in industry who want to avoid having humans do any of that work because it’s expensive. This leads to the same kind of problems any other decision to reduce quality to save money does.

Not mentioned in the article: Alaska uses ranked choice voting.

A solution for anyone confused about which similarly-named candidate to vote for is to rank the other #2.

Here is a sample ballot for anyone who would like to decide whether the names or voting process create confusion. I think it’s pretty clear that the one with (Registered Republican) Incumbent next to his name is the Republican incumbent.

That number is a bit questionable:

• It only includes sites where the server-side language can be detected. That’s often easy with PHP because of URLs ending in .php and well-known software like Wordpress which identifies itself in ways that are easy for a crawler to detect.
• It counts sites, not traffic. There are a lot of Wordpress sites out there which get three visitors a month.

The EU already did, but it lets Apple impose significant restrictions and charge fees. I don’t think it went nearly far enough.