ulyssessword

I underestimated what METR's results meant.

Specifically, they don't let you learn what the model can do and choose to do more of that. If you were deploying AI agents in a practical setting, you could simply choose not to have GPT 5.2 try to find the tallest building (3 minute task, 25% success rate), and instead do audio classification of macaques (6.0 hours, 100% success rate).

A 50% success rate could theoretically cash out at perfect reliability on half of the tasks. In practice it isn't that far off, with tasks in the 0-10% success rate band being almost as common as 10-90% ones.

Unfortunately, there are two parties involved.

The shepherd may have a 1:100 ratio of warning cost:wolf cost, but his neighbors might only be around 1:10. Giving the twentieth false alarm may be worthwhile to the shepherd, but responding to it wouldn't be worthwhile for his neighbors. And then nobody would respond to the 21st warning, even if there was a wolf.

https://aisle.com/blog/ai-cybersecurity-after-mythos-the-jagged-frontier

They can detect the problem, but not develop a working exploit. They also had a 2/3 false positive rate on a the patched version of that function.

They say that's fine because something other than the (frontier) model can do those steps, but don't demonstrate the capability anywhere I could see.

I googled the source link here: https://aisle.com/blog/ai-cybersecurity-after-mythos-the-jagged-frontier

I'm also concerned about isolating the code. It's the difference between finding a needle in a haystack, and distinguishing a needle from a single straw. Their set of models returned 12/18 false positives (and 18/18 true positives), which suggests terrible specificity to me.

I am willing to bet that despite the supererogatory behavior of these volunteers, the company and executives profited disproportionately.

Disproportionate to what?  

They probably didn't get a 4.2x income multiplier (168 hours / 40 hours), or even a 1.276x one ((2000 hours + 4 weeks * 128 extra hours per week + 40 hours extra vacation)) / 2000 hours per normal year), so I don't think it's disproportionate to the workers.

Disproportionate to society as a whole?  That would take an absurd amount of price gouging given how valuable the material was, and I feel like someone would've at least written an expose (if they weren't criminally charged).

A camera that can do facial recognition from outside of national borders doesn't need to be a petapixel one.  A mid-gigapixel camera with good optics can cover an entire city at once (or at least it could if it wasn't for all the buildings in the way).

The main barrier to petapixel cameras is that they don't serve your goal of full public monitoring (regardless of whether it's by the government or by everyone individually).

The constant angular field of view is the disagreement.  A camera in the mid-gigapixel to low-terapixel range could cover one city by using an appropriate lens at an arbitrary distance (including space).  

Any sensor finer than that would either cover substantial amounts of "boring" area (e.g. nature preserves, agricultural areas), or increase the resolution beyond your target.

I don't think you understand what a "pixel" is.

If you want to see a 1 km x 1 km area at a resolution of 0.1m, then you will need 10 000 x 10 000 = 100 000 000 points on the image, AKA 100 megapixels.  This is (ideally) independent of technology.  You can walk around and take fifty 2 MP pictures and stitch them together, you can fly a drone a few hundred meters up and take a wide-angle shot, or you can fly a satellite overhead and take a picture from space.  The distance doesn't matter.

From that, it's a simple extrapolation that it'll (again, ideally) take ... (read more)

I think a moderately-skilled person could outperform Claude here, but it's closer than you might think.  Have you thought of running this experiment with a human on the other end?  

I occasionally give technical support for industrial automation equipment, and I feel for Claude.  It's so much harder than it looks, even when you have voice+video instead of text+pictures.  

As one example of how it can go wrong, I said "Check the cables on the enclosure, and make sure they're all connected properly."  instead of "Check the three cables... (read more)

Anecdote time!  

In the past couple weeks, I:

• stopped driving my 2007 vehicle
• rented a  2025 Toyota Corolla for a while
• bought a base-model 2022 model year vehicle

The newer ones have much better handling, power, and comfort, along with some fancy features.  They also have some significant downgrades.  For the Corolla:

• The touchscreen for the radio (etc) is too bright.  It's also a touchscreen instead of buttons.
• The signal lights don't react properly.  There is no physical feedback when you lock in the switch, and it doesn't cancel

Re #3: The easiest demonstration is to do it.  How well do you think it would go over if you said "I know you like cooperative board/card games.  I found The Crew for sale at Central Gaming for $19.99, and you might enjoy playing it.  Here's their address and a $20 bill." vs. just buying it and giving to them?

Kind of against #4: A gift includes permission to have and use it.  This goes double when the recipient is a child, spouse, or anyone else the gift-giver has a stake in.  "Why do you think those binoculars are the best way to spend $1000?" Not my choice, it was a gift!  "Why do you birdwatch during our hikes?" Because I was gifted the binoculars!

If you haven't seen it in your investigations, then I doubt if raw timestamps would help.

• First, follow the link in the Youtube description to the FBI page, and download the .mp4 file.
• Next, open it in VLC (download it first, if needed), activate "Interactive Zoom" mode (Tools -> Effects and Filters -> Video Effects -> Geometry -> Interactive Zoom), set it to 800% (small slider below the left corner of the picture-in-picture) and focus on the edge of the small rooftop building.
• Last, look at 0:09-0:10.  There is clearly an object of some sort

It's visible in several frames as he walks away, otherwise it blends in with his legs.  It's also easy to mistake for another leg.

when you receive quite a few DMs asking you to bring back 4o and many of the messages are clearly written by 4o it starts to get a bit hair raising.

Am I missing something, or is that impossible?  How could it be written by 4o after 4o was taken offline (and before it was reinstated)?

At the point of death, presumably, the person whose labour is seized does not exist. I think that's a good point to consider, since I also estimate that a significant amount of resistance to the idea of no inheritance assumes the dead person's will is a moral factor after their death.

Yes, I make that assumption.  I believe I'm in very good company there, with both the general public and (many, but not all) decision theories/moral systems recognizing agreements that carry on past death.  Why would you think otherwise?

I also don't agree that you're

Inheritance is not about the children.

You ask whose labour is seized by a 100% death tax?  The parents' labour.  That's obvious enough that I feel I must be missing something.  What was your (presumably?) rhetorical question supposed to make me consider?

Inheritance is a way to get people to contribute towards prosperity for the future of the human race...by convincing them to contribute towards the prosperity of Bob, their beloved son.  Maybe you don't need a personal connection to take selfless actions, but that's not universal: I bet ... (read more)

Key paragraph:

The A-12 “practically spawned its own industrial base” (CIA 2012), and over the course of the program thousands of machinists, mechanics, fabricators, and other personnel were trained in how to work with titanium efficiently. As Lockheed gained production experience with titanium, it issued reports to the Air Force and to its vendors on production methods, and “set up training classes for machinists, a complete research facility for developing tools and procedures, and issued research contracts to competent outside vendors to develop improved

I think you're overstating your case on Science Beakers.  Take the example of titanium, as described here.  In short, what happened was:

1. Basic research happened, leading to small-scale production and basic knowledge of its properties.
2. People (including the US government) started spending science beakers on the Titanium tech node.
3. Through experience and research, they learned stuff like the fact that cadmium-coated wrenches are bad.
4. Now, we can effectively work titanium.

If it wasn't for the A-12 project (and its precursors and successors), then we sim... (read more)

What are the interactive elements?  I didn't see any, so I'm curious what the "full experience" was supposed to be.