ianmcgowan

That's the horror..

It is, apocryphally, a gotcha test for Gen-alpha. As is being able to return the correct change or make a phone call to a human being. Reminds me of the iRobot line: "Can you create a symphony?", "No. Can you?".

Now I'm watching this thread for the consultant hired by the film to show up and explain why each of those goofs was caused by the director explicitly asking for them...

It's fascinating the extent to which all the models rely on text - it's like they have severe (but not total) aphantasia:

"We hypothesize that this phenomenon emerges predominantly from a misassumption about how these systems are trained. Modern multimodal models are developed on web-scale corpora and are commonly built on top of pretrained large language models, which makes them extraordinarily strong at language modeling, retrieval of statistical regularities, and reconstruction of likely contexts from sparse cues.[48, 25, 24] During the multimodal training, the models are presented with the image, a textual question, and are expected to reconstruct the correct answer. Lacking access to an entire text corpora, a human would intuitively answer the question based on the image in that setup; but we should not infer that this would be the default approach for an AI model. Incentivized to generate the correct next tokens, models might learn to easily ignore the visual information and rely only on their vast prior knowledge, taking the shortest route to the correct answer.[36, 5, 48]"

The crazy thing is that based just on the text in the questions a model was able to "guess" answers:

"When fine-tuned on the public training set of this dataset with images removed (i.e., trained in mirage-mode), our 3-billion-parameter, text-only super-guesser outperformed all frontier multimodal models, including those exceeding hundreds of billions of parameters, on the held-out test benchmark (Figure 3c). It also surpassed human radiologists by more than 10% on average, relying entirely on hidden textual cues in the questions and the structural patterns of the benchmark. In addition, our super-guesser was able to create reasoning traces comparable to, and in some cases indistinguishable from, those of the ground-truth or those generated by frontier multi-modal AI models."

Surely "Player Piano" is more relevant?

Or Roko's Basilisk

Probably the burning man essay, which is one of the best things I've ever read online.

https://open.substack.com/pub/samkriss/p/numb-at-burning-man

A protest demonstrates a level of unhappiness with a group or policy. People may not believe what they see on the news, facebook, or youtube, but hopefully we have not reached a point where they refuse to believe what they see with their own eyes.

The point is to demonstrate "we are not alone in this feeling", that's it...

Sounds like the plot of "Big Brother" by Cory Doctorow

Try "No agenda, no attenda" - see if a touch of levity helps.

Lovely essay, tone reminds me this book which has a similar vibe.

https://www.thriftbooks.com/w/truck-on-rebuilding-a-worn-out...

Kernighan!

That's exactly what the Lizard overlords are doing! /s - wouldn't be surprised there's a few conspiracy theorists who believe this...

Parent comment is working from The Gervais Principle[1] of office politics. There are companies where that's an accurate description of reality, unpleasant though it might be.

[1] https://www.ribbonfarm.com/2009/10/07/the-gervais-principle-...

This is where defensively blocking off your calendar comes in!

I don’t get why municipal investment isn’t more of a thing - why can’t cities and counties be their own PE, buy a small fire truck company and turn it into a non-profit? If they captured the surplus that’s going to the PE company surely costs would be cheaper?

I guess that’s socialism though, so not gonna happen.

A tangent - sounds like https://en.wikipedia.org/wiki/The_Book_of_Koli - a key plot component is a chatty Sony AI music player. A little YA, but a fun read..

Popper and "the paradox of tolerance" to the rescue. You can, and should, tolerate anything but intolerance.

A chromebox mounted behind the monitor did the trick for me. Haven't had an emergency wipe/reinstall in years. Also, a tablet with keyboard takes some of the pressure off having a "computer" and you can go with iOS or Android depending on what phone they use.

I moved from IC to manager to director. Meetings in general are a huge time suck, and by the end of my tenure, I was basically triple-booked for the entire day every day. Every evening was an exercise in deciding who to piss off. And the emails! <Marvin/Android Voice>Don't get me started on the emails! :-) Finishing the twilight of my career as an IC consulting, and it's much better for me.

In California, as a non-exempt employee (basically not a manager if you're at a big company), you'd have to be paid for that on-call time with those requirements. The key term is "restricted" and the 10 minute expectation is quite a severe restriction.

https://www.dir.ca.gov/dlse/callbackandstandbytime.pdf https://www.shrm.org/topics-tools/tools/policies/california-...

This doesn't apply to a lot of people reading, but just a PSA for those in CA where it might.

I moved to "the city" in 1989 from England, and people were complaining then about yuppies and it wasn't the same as the good ol' days of the 60's.

SF seems to be a lot more in-flux compared to other cities, so if you don't like the scene now just wait a few years and a new one will be along :-)

That doesn't explain why math is so amazingly effective at describing reality - all the pure mathematical results seem to eventually end up being useful at describing some new aspect of the world (to the chagrin of the pure mathematician).

Maybe it's the other way around? Our minds work this way because that's the rules and we're just emergent from that reality. It's hard to argue that symmetry is not a mathematical ideal first, and a biological approximation to that ideal second.

I love the mental picture from Greg Egan's Diaspora of the truth mines, literally hacking the math like chunks of coal:

"If ve ever wanted to be a miner in vis own right -- making and testing vis own conjectures at the coal face, like Gauss and Euler, Riemann and Levi-Civita, deRham and Cartan, Radiya and Blanca -- then Yatima knew there were no shortcuts, no alternatives to exploring the Mines firsthand. Ve couldn't hope to strike out in a fresh direction, a route no one had ever chosen before, without a new take on the old results. Only once ve'd constructed vis own map of the Mines -- idiosyncratically crumpled and stained, adorned and annotated like no one else's -- could ve begin to guess where the next rich vein of undiscovered truths lay buried."

A subsidiary of Lumon? Do you have to be severed to work in MDR?

Let's hope they don't want to paint the bike shed!

I always start with "SELECT COUNT() FROM ..." for this reason, and then go back and replace the COUNT() when I get a handle on the tables/filters/groups involved..

Chesterton's point wasn't that the fence should never move (which is what you seem to be implying), but that it should only be moved (or technically removed) when you understand why it was initially put in place.

I'm sympathetic to the idea that there is bloat and over-regulation, but most of the laws and regulations on the books are in response to a bad thing happening. It's kind of like a legacy code base - just deleting the repo and starting over from scratch is usually not the best idea, it takes some careful refactoring and judicious tests to move in the right direction.

It's sort of an interesting idea - what are "tests" in the context of the legal/regulatory framework? The constitution? The judiciary?

Yes! I do this when doing anything destructive ("rm -rf dir", "drop db", etc.) - i just stop and say out loud what environment I'm in, what the command I'm running is going to do and why I am doing it. This is in a remote-work situation, in an office I would have someone come double check the really destructive things before hitting enter..

Really puts the auto- in didact! Very curious to hear how this worked for you; it’s almost directly the opposite of the copilot approach.

I learned assembler by typing in listings from magazines and hand dis-assembling and debugging on paper. Your approach seems similar in spirit, but who has the times these days?