hrokr

At the end of the article, it says OpenAI prioritizes work in service of three north stars: navigating the next period of AI progress, delivering the benefits, and empowering everyone.

Strangely, it doesn't mention not wrecking humanity or the planet. Considering how so many key figures have called this out, you would think they would at least give lip service to the topic.

Sycophancy As A Service

It's interesting that the increase in Vim commits is concurrent with the fist release of NeoVim. They've certainly both been good for each other and I'm hoping this won't change with Bram's passing.

If you have 1 bad officer but 999 who go along with it, you have 1000 bad officers.

And remember, this is just one case.

"Instead of making headlines with reform, though, the TSA has become better-known for stealing money from travelers. In September, news reports called out the case of TSA working with other agencies to seize $27,600 from a Texas man, apparently because he was traveling to Oregon where marijuana is legal in conflict with federal law; he was never charged with a crime. Last year, the Institute for Justice reported that TSA and its sister agencies at Homeland Security "seized over $2 billion in currency at airports" between 2000 and 2016."

https://reason.com/2021/11/19/after-20-years-of-failure-kill...

At the time the TSA was hired the majority of workers had previously been airport security which is was low-skill/low-pay job. When Bush Jr established the TSA, those same people were federalized. Meaning Federal pay, benefits, retirement, etc for the same low skill level.

They haven't improved much since then. The organization has essentially made Security Theater an event with audience participation. Yet they routinely miss threats -- like they have a 95% FAILURE rate at points.

https://www.schneier.com/blog/archives/2015/06/reassessing_a...

Incidentally, that post is fairly well known amongst the security minded as you see here.

https://reason.com/2021/11/19/after-20-years-of-failure-kill...

I know a trauma therapist. This is what she says:

If you're talking real trauma, I'd seek out a therapist licensed in EMD or EMDR. Data shows it's faster and more transformative that Cognitive Behavioral Therapy. It seems to work well for critical past events (rape, driving accident, etc). Sub-clinical trauma (like that often experienced by police officers) seems to respond if the event can be crystallized.

If you're talking about the trauma of something that isn't to that scale, say the trauma of being dissatisfied with how life is turning out, consider meditation.

I agree you'll definitely want to read Elements of Statistical Learning but there are a few more, namely Think Stats and Think Bayes.

Since no one has really said much about Bayes yet, I think it worth mentioning just how useful it is in DS and ML. A Bayesian approach makes a very good baseline and often one that is hard to beat.

If you're not particular fluent with Probability and Statistics now, let me suggest you add in Khan Academy (make sure to pick the CLEP version) and JBstatistics. Khan has the advantage of quizzes (so you're not just kidding yourself that you know the material). JBstatistics has the advantage of really good explanations. You'll probably want to watch Khan at x1.5 speed.

I think they’re confusing pesticides with herbicides.