ajcp

My go-to SSID will always be "Robert'); DROP TABLE Students;--". For all others I just use a ship name from The Culture universe[0] like "Of Course I Still Love You" or "Just Read the Instructions".

0. https://theculture.fandom.com/wiki/List_of_spacecraft

Having watched MSFT slowly chip away at their traditional bread-and-butter OS model with things like OneDrive and Office in the browser, Azure and then WSL, and listening to the Acquired podcast episodes on Microsoft, I wonder why they haven't simply released a Microsoft Linux by now, if only out of pride? Do they feel that by doing so they're broadcasting that they're no longer a computing philosophy leader, and merely a market preference fulfiller (which is itself a backhanded way of saying they meet market demand I guess).

-> and make it happen.

I think OPs point is they failed on this part. "Making it happen" should have been ensuring a compliant and approved version of the software was the one made available to the developers. At a large scale that is done via device management, but even at a medium sized enterprise that should have been done via a source management portal of some sort.

I'd be interested to know why buying, installing, jerryrigging, and (presumably every time you did a load of laundry) hooking and unhooking collapsible hosing for a washer and dryer in a bedroom you worked from, was in any way more convenient or cheaper or useful than just using the communal laundry room or a dedicated laundry service?

I'm hosting a VSCode server with Termux/Ubuntu container on my old Pixel 6a and I cannot overstate how awesome it is for just a fun dev setup, especially with a tablet. Easy to nuke and start clean too!

Very useful info; much appreciated!

The only reason I haven't gone over to Linux is gaming with my RTX card. Interested to know your gaming setup and distro. Any stability/compatibility issues?

I agree with this, especially in the case of camera feeds that are run by organizations that are supposedly servicing the public.

That being said I also don't wonder if there is a point where we're just crowdsourcing the police state?

->...it's commissioning.

I like this. I'm going to see if my boss will go for me changing my title from Solutions Architect to Solutions Commissioner. I'll insist people refer to me as "Commissioner ajcp"

I have a pretty deep "smart home" setup and it's all run locally from a laptop in my closet with Home Assistant OS. I have run into 0 limitations. All my devices are kept on their own dedicated Zigbee mesh and/or network separate from my LAN. Only way to communicate in or out is via Tailscale. It's incredibly easy to get started too.

Given TikToks insane creator adoption rate is Meta developing these models to build out a content creation platform to compete?

I highly recommend the book; it definitely completes the series well.

FWIW there are actually 4 books in the Three-Body Problem "trilogy". The Redemption of Time was written by a fan who felt the series didn't provide closure and was recognized as canon by Cixin Liu.

No, I agree, I just wanted to call it out.

-> Rich developer spends $15k to run a model slightly faster.

I love these and I know this is all in good fun, but I feel like this one is a little unfair to Jeff. He's a content creator and he didn't actually buy the rig. If he's rich it's because he creates content like this.

Having known nothing of this field before now I have to say your excitement has me excited!

This is very compelling, very nice work!

I really would like to know how good this would be for a corporate Internal Audit workflow/professional.

And so the Kākāpō Benchmark was born

It's sadly all too common. I worked at a Fortune50 retailer with a massive IT org. Was on a call one day with one of our most Senior Ent Arch who was excitedly telling me about how "these Java scripts" were the hot new thing and we were building "modern web 3.0 pages" with them. He did not understand the difference between Java and JavaScript or a blockchain branding exercise and SPAs.

-> They were concerned about the increased attack surface resulting from including the current 100K+ lines C++ libjxl reference decoder, even though most of those lines are testing code.

Seems like Google has created a memory-safe decoder for it in Rust or something.

"I'll get you to the moon safely" = NASA "I'll get you to the moon" = me with a wood chipper and a very powerful potato gun The startup charged for "AI note- taking", not simply "note taking".

As long as the ergonomics with the SDK stay the same. Jumping up to a new model this far in is something I don't want to contemplate wrestling with honestly. When we were forced off of 1.5 to 2.0 we found that our context strategy had to be completely reworked to recover and see better returns.

-> put Gemini 2.5 at the top of the pack

I have come to the same conclusion having built a workflow that has seen 10 million+ non-standardized PDFs (freight bill of ladings) with running evaluations, as well as against the initial "ground-truth" dataset of 1,000 PDFs.

Humans: ~65% accurate

Gemini 1.5: ~72% accurate

Gemini 2.0: ~88% accurate

Gemini 2.5: ~92%* accurate

*Funny enough we were getting a consistent 2% improvement with 2.5 over 2.0 (90% versus 88%) until as a lark we decided to just copy the same prompt 10x. Squeezed 2% more out of that one :D

It is but I feel it's main value prop as a developer friendly abstraction layer has been very well solved for by the actual model providers themselves, while LangChain itself have become more bloated, clunky, and to under-opinionated.

man, I can't tell you how much that means to hear! Glad you liked it. Check out the console for more :)

Of course!

That's the move right there.

-> I would be so entertained if I found out an AI lab had wasted their time cheating on my dumb benchmark!

Que intro: "The gang wastes their time cheating on a dumb benchmark"

NotebookLM came out of Google Labs though, and in collaboration with outside stakeholders. I'm not sure I would call it a success of "bottom-up" culture, but a well realized idea from a dedicated incubator. That doesn't necessarily mean the rest of the company is so empowered or product oriented.