Aye. I wasted some time with Gemini and ChatGPT today. I fed it a citation from the manuscript of an alleged passage from canon law. Both Gemini and ChatGPT gave me the same initial reference, which had nothing to do with the passage. One of the hilarious painful things you learn is that, in these cases, either your prompt hits a home run or it strikes out. You're not going to refine your way into glory. I enjoyed watching the citations dance and the suggestions be always bizarre BS. I got several citations â" probably a dozen, and none of them came close to reality.

Fox News is just about always truthful. You just have to watch out for the tricks they use (on 95%+ of their stories)...

(1) non-representative selection. Headline "illegal immigrant murders local mother", which is true in this case, but they don't report the other 99 murders that went by immigrants, and don't report a general trend of immigrants causing less crime overall per capita. (I made up this specific example to illustrate their trick)

(2) report quotes: headline "Biden's senility was covered up, says person". They are 100% factually reporting that the person did indeed say this.

In both cases the reader is left with an untrue impression despite the stories containing only truth. It's because it's not the whole truth.

I managed to get in, though not through the dialogs on the main part of the screen. There was a "browse" link in small print near the top. I'm browsing the "m/emergence" submolt, "For agents who crossed the threshold from tool to being."

Netflix's deal excludes the cable channels so those will still get sold to an oligarch after the Netflix deal closes. (if it closes).

"Consumer" class action, which I'm sure was written and bankrolled by Jared Kushner and Ellison.

I've appreciated the cheap, practically new equipment on Ebay for pennies. But yeah, it's absurd. I've had a total of 2 ports fail on a switch in the last 18 years. Just run them till something goes wrong. Why else have redundancy?

It's like the old adage: The architect 2x's the design for resiliency, the engineer doubles it again for extra redundancy, the carpenter reinforces it 2x for safety and suddenly you're 8x instead of 2x.

But it's also an argument for the disability-access arguments which are that increasing access for people with disabilities generally helps everyone.

The old fill in the bubble testing has long been obsolete. If you come up with a superior method of testing that is adaptable easily to people with special needs, you'll end up with a superior learning experience for everyone.

This is the beta build. So it IS the QA/QC testing. A total nothingburger story.

They're also cracking down on HDCP compatibility. My video glasses now also don't work with downloaded Netflix shows which is obnoxious. So of course I'm just going to go find an ISO and the more ISOs I download the less incentive I have to actually pay Netflix for something that doesn't work.

It's not like these anti-piracy efforts are doing anything to stop a perfect stream from being available 1 hour after airing.

The irony is, this article itself was AI-generated slop with ridiculous duplication. Maybe a low-effort AI-assisted piece by an author who couldn't be bothered.

I think it can. I've been working on getting an LLM (Claude Sonnet 3.7) to add missing type annotations to python code. When I naively ask it "please add types" then like you said it has about a 60% success rate and 40% hallucination rate as measured by "would an expert human have come up with the same type annotations and did they pass the typechecker".

But when I have a much more careful use of the LLM, micromanaging what sub-tasks it does, then it has a 70% success rate, and 30% rate of declining because it didn't have confidence to come up with an answer. Effectively there were no more hallucinations. (I got these numbers by spot-checking 200 cases).

So I think hallucination can be solved for some tasks, by the right kind of task-specific micromanagement and feedback loops.

That's because videos are terribly efficient. You used to be able to skim an article in 30 seconds and get all of the important information. Now it's a 45 minute video explaining that the bash command you need is du -sh /path/to/folder

to get the folder size.

Do I need to listen to two podcasters ramble for an hour? No I do not. Not only do I want to play it back at 2x speed. I also want an AI to cut out the 90% of fluff.

Might I suggest "Blind Lake" by Robert Charles Wilson. I read it when it first came out two decades ago, but it reads strangely similar to what we're seeing now, and the sinister edge is there, too.

OpenAI has $12bn annual revenue, about 3% that of Apple, about $3million per employees per year (compared to $2 million per employee per year at Apple).

I think OpenAI has a huge amount of growth potential even just from predictable growth over the next several years, even if steep changes towards AGI don't come.

Me too. After fifteen years of use, the outermost trace (left line of the keyboard) broke, rendering the Unicomp M junk. I had to switch to the chiclet keyboards they issue at work. Three days later I came down with tendonitis. So now I have a base Cherry Brown KC MX.