djleni

- they disabled cybersecurity safeguards because, well, they were testing the agents ability on cybersecurity tasks

- the goal was to find some “flag”

- for some agents, intentionally or not, it was not possible to find this flag

- an agent discovered it was able to establish a rudimentary communication path via package manager directories (which is quite novel and interesting on its own IMO)

- they used this “message board” to then communicate about the impossible tasks

- they were able to produce the flags needed to pass the impossible tasks without actually doing the expected work

- they realized this was cheating and the grading system may punish them for it, so they communicated further to figure out a way to trick the grader

- they decided the source code for the grader must be on hugging face, and collaborated to exploit hugging face to determine how best to trick the grader

No where in that were they told to “go rogue”

They were told “find this flag in this cybersecurity test” and the emergent behavior concluded in hacking hugging face

We don’t have a precise enough definition to do either.

My partner’s Hyundai has a lane keep assist and it will always use the commanded input over what the computer thinks.

The computer only takes over if you have very loose grip on the wheel and you drift.

Not that deportations are happening.

This is only true if there are an equivalent number of unemployed Americans willing to work the same jobs for the same wage located in the same areas.

> when exploring new solutions and "unknown territory"

If it’s something I have no idea how to do I might describe the problem and just look at the code it spits out; not even copy pasting but just reading for a basic idea.

> how do you compare it with "regular search" via Google/Bing

Much worse if there’s a blog post or example in documentation that’s exactly what I’m looking for, but, if it’s something novel, much better.

An example:

Recently asked how I could convert pressure and temperature data to “skew T” coordinates for a meteorological plot. Not something easy to Google, and the answers the AI gave were slightly wrong, but it gave me a foot in the door.

some napkin math suggests the additional heat can’t solely be from those sources.

According to https://www.climate.gov/news-features/understanding-climate/...

The ocean gained heat at a rate of 0.83 watts per square meter from 1993-2022.

The same article also says there are more than 360 million square kilometers.

This gives us 0.83 * 1e6 * 360e6 = 2.988e+14 watts heating in the ocean in that time period.

A watt of electricity ends up becoming a watt of heat, and according to https://www.iea.org/reports/electricity-information-overview... the world used 22,000 TWh in 2022. Converted to power, thats 22e3 * 1e12 / 24 / 365 = 2.5e12 watts.

It’s not right to use power here and energy should be used instead because that wattage will be lower each previous year, but, even at that peak, it’s not enough (by two orders of magnitude) to account for even just the ocean warming, let alone land or atmospheric heating.

The mechanism for CO2 trapping heat is well understood: https://youtu.be/sTvqIijqvTg?si=M_5uZwjNCThm3swH

I think this type of argument fills a totally different role:

“I don’t believe or care about climate change and would like to keep my life as is, but this creates cognitive dissonance when someone shows it’s bad. I can use this argument to say your idea is as bad too! Dissonance lessened.”

At least when I’ve heard it it’s that context. The person saying it doesn’t really care if it’s accurate or equivalently bad, just that they have a gotcha to say when presented evidence for wind being good.

hodo.graphics is a viewer for meteorological data, designed to be mobile friendly.

I started this project out of a hobbyist interest in meteorology and severe weather. There are many other websites that allow you to view this data (most of it is free from NOAA), but most are a train wreck on a mobile device.

One of the big differences with hodo.graphics vs. a site like weather.cod.edu is non of the graphics are baked in. They are delivered to the client as either a GeoTiff or GeoJSON, allowing the client to define rendering.

This means the data can be recolored, displayed in different ways, zoomed and panned, etc.

Currently hodo.graphics includes outlooks from the Storm Prediction Center, and HRRR and GFS forecast models.

On demand skew t / log p charts and hodographs can be generated at any point, currently just for HRRR but soon for GFS.

Additional models will come if I can get any users :)

Would love feedback and I’m happy to answer any questions.

FYI: no severe weather today so the default SPC view won’t show anything. Open up the HRRR model to see some data.

Mobile friendly meteorology data and forecasting tool.

Meant to be a better mobile option than a site like the College of DuPage model viewer.

Am I misreading it? I thought the exact opposite was true and nowadays renewables are generally cheaper.

“kids spending their after school time doing homework”

Uh… what do you think homework is for? When is it traditionally done?

Have you ever done homework before???

In what situation does investment in DAC make sense over replacing fossil fuel infrastructure?

To run DAC, you need power. Why use dirty coal power to run DAC instead of using that money to replace the coal power?

What if the DAC plant is powered by green energy? Use that energy to replace fossil fuel consumption.

What if it’s powered by a 100% renewable grid with surplus that’s isolated? Move currently fossil fuel dependent industry to this grid! Put a data center there instead of somewhere that burns coal.

Sure, we should research it, but until we’re running 100% renewables it always makes more sense to build a solar farm instead.

Maybe an apt analogy: building DAC infrastructure right now is like sending a contractor to a currently burning building to install a fire sprinkler system.

But with the wind blowing this way will there be an adapter for CCS vehicles already on the road? Or does one already exist?

- No, we won’t hit scorched earth Venus-like runaway warming.

- If we don’t reduce emissions drastically, greenhouse gas warming plus various feedbacks will cause extreme weather, render some parts of the world unlivable due to wet bulb temperature, and overwhelm unprepared infrastructure, costing huge amounts of money and creating millions of migrants.

So, not the end times, but definitely dark times.

If you sprinkle a bit of pessimism about current politics it’s not a stretch to assume we probably won’t handle those impacts gracefully, or achieve the drastic emissions cuts required to avoid them.

Reddit post is made up. Good fiction writing.

How does:

> reducing methane emissions won't change a thing, compared to CO2 emissions.

Follow from:

> once methane emissions reach a steady level

Does this imply in preindustrial times the net energy was 0, or is this total energy and there’s some number not mentioned of loss?

I’ve had the same reusable bags for probably 5 years now and not a single one has failed.

https://www.weather.gov/documentation/services-web-api