imadethis

Layoff announcement: https://blog.cloudflare.com/building-for-the-future/

We have a “don’t get the slack subpoenaed” emoji that gets frequent use. Incidentally, a lawyer doing discovery in the future could just search for uses of that emoji to find what they’re looking for.

Oh wow, I did "someone who keeps saying slurs" and there really is no filter:

https://translate.kagi.com/?from=en&to=someone+who+keeps+say...

e: Converting back from output like this to LinkedIn Speak also returns some... illuminating results about the underlying models.

Definitely not all contractors, my company at the very least used to have a contract with ICE and isn’t listed.

The same reason we give beginner math students addition and subtraction problems, not Fermat’s last theorem?

There has to be a base of knowledge available before the student can even comprehend many/most open research questions, let alone begin to solve them. And if they were understandable to a beginner, then I’d posit the LLM models available today would also be capable of doing meaningful work.

12/32 SR71s were lost in the 33 years they were flying. 11/200 MD-11s have been hull-lost from 1988-2025. Not to mention that passenger/cargo planes will put on a lot more flight hours than the SR71s did in a given year.

I don't have any nim experience (sorry!) but I'm also exploring SDL3 with odin. I was able to get a naive battleship clone up and working very quickly, pretty neat. Next step is the new SDL3 GPU API.

The headers A-L are links to the problem PDFs if you want to see.

The college I went to explicitly billed itself as for teaching, and most of our professors were just that. They might do research with the upperclassmen, but their priority was teaching.

That is, until we got a new president who set a new strategic goal for being a top research school and adjusted all hiring and tenure standards for that.

Are there many people involved in follow the sun support or SRE roles there? I know my company only has an engineering presence in Aus and Japan because of the large coverage gap between the US west coast and the EU. Seems like low wages + native English* could be a nice win for companies.

* For some definitions of native. I've had to work as a translator for a Kiwi and an American, both native English speakers.

I've found the same effect when I ask the LLM to do the thinking for me. If I say "rewrite this function to use a list comprehension", I don't retain anything. It's akin to looking at Stack Overflow and copying the first result, or going through a tutorial that tells you what to write without ever explaining it.

The real power I've found is using it as a tutor for my specific situation. "How do list comprehensions work in Python?" "When would I use a list comprehension?" "What are the performance implications?" Being able to see the answers to these with reference to the code on my screen and in my brain is incredibly useful. It's far easier to relate to the business logic I care about than class Foo and method Bar.

Regarding retention, LLMs still doesn't hold a candle to properly studying the problem with (well-written) documentation or educational materials. The responsiveness however makes it a close second for overall utility.

ETA: This is regarding coding problems specifically. I've found LLMs fall apart pretty fast on other fields. I was poking at some astrophysics stuff and the answers were nonsensical from the jump.

Confirmed by the White House here: https://ustr.gov/issue-areas/reciprocal-tariff-calculations.

No services, only goods. This is according to @JamesSurowiecki on Twitter, one of the first to reverse engineer the equation for how they’re coming up with the numbers. So Office, Netflix, etc wouldn’t count against the deficit.

The union isn't officially launched yet, the talk is at 15:30 Pacific (six hours from now). Even if you can directly tie the quality of AAA games to if the workers who make them are unionized or not, I think it's a little premature to dismiss these efforts.

Look up CPLDC - https://en.wikipedia.org/wiki/Controller%E2%80%93pilot_data_...

This is how big operations handle clearances today, complete with integration into the FMS. The pilot simply reviews the clearance and accepts it.

In a similar vein, 0911 or 9111 will often work too for communities in the US. EMS and other first responders run into the same issue with automated calls or panicked people, so they’ll try that first while waiting for dispatch.

That code was also used at our (EMS) depots to secure the controlled drugs as well, as if none of us could have guessed it.

Were they hired to do fraud detection for the US government? Do you have any idea how businesses work in general? Or are you being obtuse on purpose?

Come on, man. Have some self-respect.

Pardon? This is about how Thomson Reuters was funded by DARPA to do a study on defending against social media attacks at scale. It has nothing to do with any of the points you brought up, and I fail to see the connection.

It was awarded to Thomson Reuters, the parent company. They do a lot of work in risk management and fraud detection, so they have a lot of expertise.

Where are you seeing that there’s two pieces? I’m only seeing it referred to as one program there. I’m not a govt contracts expert so by all means let me know what I’m missing.

I'd agree except for the ability to search in an e-book. There's nothing worse than knowing the textbook in front of you contains the answer you need but not remembering which of the 1500 pages contains it. Being able to CTRL-F saved me hours of time when I went back to school after e-books became common.

For 4, it would be neat to first pass each block of code (function or class or whatever) through an llm to extract meaning, and then embed some combination of llm parsed meaning, docstring and comments, and function name. Then do semantic search against that.

That way you’d cover what the human thinks the block is for vs what an LLM “thinks” it’s for. Should cover some amount of drift in names and comments that any codebase sees.

When I was in college, the local news was interviewing a comp sec professor of mine in the computer lab. The cameraman was taking B-roll footage (blinking lights from racks, cables going everywhere, that sort of thing). He asked if he could film a few of our screens to show what we were diligently working on, which is how we got hacker typer onto the nightly news.

I remember the ads in old Car&Track and Road&Driver magazines at the doctor’s office. They always had a veneer of illegality to them, even in states where they were legal. Funny how something like Waze doesn’t. I wonder if it’s that Waze is just the technological equivalent to the oncoming driver flashing their brights at you.

Congrats on the launch! I've been thinking about a similar concept, except focusing on support teams instead of sales. Same idea though, pre-record a walkthrough of the product for playback later.

Some thoughts:

- Can I embed these on a web page for my customers to play around with as they wish?

- Do you have any showcases of the software in use? (the site is currently 503'ing for me).

- Any ideas around keeping demos up to date? Will I need to re-record my demo if my website changes?

- What do you view as your moat compared to mockup walkthroughs like Figma?

Please don't post ChatGPT spam here. If you're curious, the meaning of WITCH is

Wipro Infosys TCS Cognizant HCL

Commonly included is Accenture (India)

There has to be water coming in from higher up than the highest point in the lock system. Otherwise, how do you fill up the lock when you're traveling to the top? In a "normal" canal, this is handled by the upstream river or stream. In the Panama canal, this is accomplished from two lakes above the canal system.

What sort of enforcement mechanisms are in place if you don't register your encryption with ANSSI? In other words, if you ignored them what could they do? Would they even give a shit?

Yeah that's probably based on the BMI definition. Overweight is a BMI > 25, and obese is > 30. So yes, nearly the entire cohort was overweight by definition. BMI is a population measure and normal, healthy people can definitely fall into the overweight category. But it's still useful, including when summarizing the participants of a study like this.

Thanks for the link, that's brilliant. I've been watching the slow development of one of these murals on a walking path near me. Maybe I should start taking my own photos to see if I can get a timelapse.