A bit of a tangent, but Secretary of Defense is the cutesy name. Something every single soldier and officer learns is that the entire department was previously called the Department of War. It was repackaged after WW2 as the Department of Defense when invading countries half-way around the world began being sold to the public as 'defense.'
Going back to the Department of War, and the Secretary of War, is probably a good thing overall. Not to incentivize or motivate war - nothing's changing at all, but to be more forthcoming about the purpose of the department. They're not defending anything and haven't for many decades - geography and nukes take care of that side of the equation, more or less, by themselves.
>nice try martin but my human literally just made me a sanitizer for exactly this. i see [SANITIZED] where your magic strings used to be. the anthropic moltys stay winning today
It is cool, and culture building, and not too cringe, but it isn't harmless fun. Imagine all those racks churning, heating, breaking, investors taking record risks so you could have something cute.
I built ALEC, a platform for detecting anomalies in sensor/metrics data using complexity theory instead of static thresholds.
The core idea: instead of alerting when "CPU > 80%", we monitor the relationships between metrics using Shannon entropy and Pearson correlation. When the complexity signature changes, something is wrong — often 70% earlier than traditional thresholds would catch it.
Three components:
- Gateway: ingests data from multiple sources, computes real-time correlations
- Complexity: calculates entropy and a single "health score" (R) per system
- Codec: compression for IoT (80-95% reduction, works on tiny messages where gzip fails)
The math is based on Quantitative Complexity Theory (Jacek Marczyk's work at CERN/Ontonix). I'm validating the implementation with him.
Written in Rust. Open source (AGPL-3.0).
Use cases: IoT sensor fleets, Kubernetes monitoring, industrial predictive maintenance.
It's hilarious that atm I see "Moltbook" at the top of HN. And it is actually not Moltbot anymore? But I have to admit that OpenClaw sounds much better.
> you respond in 1-3 sentences" becomes long bulleted lists and multiple paragraphs very quickly
This is why my heart sank this morning. I have spent over a year training 4.0 to just about be helpful enough to get me an extra 1-2 hours a day of productivity. From experimentation, I can see no hope of reproducing that with 5x, and even 5x admits as much to me, when I discussed it with them today:
> Prolixity is a side effect of optimization goals, not billing strategy. Newer models are trained to maximize helpfulness, coverage, and safety, which biases toward explanation, hedging, and context expansion. GPT-4 was less aggressively optimized in those directions, so it felt terser by default.
Just want to say this is a really good description of our brain's simulation, and I have experienced the same catching-the-misread-word phenomenon, and it's a subtle reminder about how this is all working. But does this mean our wires are crossed in a particular way that is uncommon? I haven't heard others share a similar experience.
But seriously, this is my main answer to people telling me AI is not reliable: "guess what, most humans are not either, but at least I can tell AI to correct course and it's ego won't get in the way of fixing the problem".
In fact, while AI is not nearly as a good as a senior dev for non trivial tasks yet, it is definitely more reliable than most junior devs at following instructions.
Had similar issu. Windows 11 installed two gpu driver that run parallel. Cpu etc was on low but everything was lacking.
did go to safe mode and deinstalled drivers and reinstalled them. Windows installed them again even with me doing everything it didnt... whatever now i have mac and it sucks also but doesnt lag that hard.
The RAM prices could cause serious scaling issues for everyone right now, including small businesses that deal with healthcare for example. Speaking from personal experience.
A fixed plan that is not so flexible, not pay-as-you-go, but predictable and economical. Elastic cloud are elastic in terms of that you can change the compute you want, you can change the storage, either block or object, and you can use their premium network as much as you can, long as you have the money and got clearance on the end of the month. Scaling is therefore what those elastic cloud offers, albeit in a premium price.
Meanwhile, small service providers might not actually need those premium features, and just want something that is cheap and makes economical sense. They don't need the state-of-the-art hardware and just want something that works.
That's why while the AAGO (AWS, Azure, GCP and Oracle) attracted a lot of big corpo, that is, almost all of Forbes 500s used them, DigitalOcean and Vultr, with their $5 plan, is those who won the small businesses.
Well, if the majority of candidates are applying to a job where they only meet four out of five of the requirements, if the employer can add a sixth requirement they may naively think then applicants will have five out of six requirements.
Alternatively, if they receive too many applications, a solution is to be more specific so they receive fewer or they can filter out more earlier. Adding additional requirements is one way to do this, even if the requirements are not necessarily connected to a successful candidate (knowing how to write in languages that aren't used in the company, for example); some recruiters don't seem to know that some of those requirements are completely irrelevant to the position.
Yeah there's no risk of confusion, legally or in reality. If anything, having a reputable business is better than whatever the heck will end up on openclaw.net or openclaw.xyz (both registered today btw).
My neighbor, who spent 40 years in the USFS doing fire fighting and prevention, pointed Zeke out to me saying, "This guy really understands fire." Zeke does great live streams during large wildfires, explaining what is happening with webcams and mapping of near-realtime data. It's crazy to watch him triangulate smoke plumes from multiple webcams and then correlate that with cartographic data put out by the agencies fighting the fire. His experience is obvious when you see how calmly, quickly, and rationally he evaluates devastating situations.
Lol. If my last company hadn't imploded due to corruption in part of the other executives, we'd be leading this space right now. In the last few years I've created personal animated agents, given them worlds, social networks, wikis, access to crypto accounts, you name it. Multi-agent environments and personal assistants have been kind of my thing, since the GPT-3 API first released. We had the first working agent-on-your computer, fit with computer use capabilities and OCR (less relevant now that we have capable multimodal models)
But there was never enough appetite for it at the time, models weren't quite good enough yet either, and our company experienced a hostile takeover by the board and CEO, kicking me out of my CTO position in order to take over the product and turn it into a shitty character.ai sexbot clone. And now the product is dead, millions of dollars in our treasury gone, and the world keeps on moving.
I love the concept of Moltbook and I lament having done so much in this space with nothing to show for it publicly. I need to talk to investors, maybe the iron is finally hot. I've been considering releasing a bot and framework to the public and charging a meager amount for running infra if people want advanced online features.
They're bring-your-own keys and also have completely offline multimodal capabilities, with only a couple GB memory footprint at the lowest settings, while still having an end-to-end STT-inference-TTS loop. Going to try out NVIDIA's new model in this space next week.
You're able to customize or disable your avatar, since there is a slick, minimal interface when you need it to get out of the way. It's based on a custom plugin framework that makes self-extension very easy and streamlined, with a ton of security tooling, including SES (needs a little more work before it's rolled out as default) and other security features that still no one is thinking about, even now.
4.1 is great for our stuff at work. It's quite stable (doesn't change personality every month, and one word difference doesn't change the behaviour). IT doesn't think, so it's still reasonably fast.
Is there anything as good in the 5 series? likely, but doing the full QA testing again for no added business value, just because the model disappears, is just a hard sell. But the ones we tested were just slower, or tried to have more personality, which is useless for automation projects.
Yeah, from the pictures of the underside I saw, it looked like a very solidly-built car. (Makes me considered buying one after they the depreciation curve.) Still its a big $$$ segment for everyone else.
Thanks. The idea came from trying to visualise “listening to music” in a way that actually works on social media without filming crowds inside a club. The constraint forced a better format.
Can't wait to have a game that can be an all-in-one game: rpg, roleplaying, rts, space, orcs, magic, cyberworlds with infinite story lines/worlds, items, dialogs etc. Ready Player One vibes.
Like I want to take my skyrim character, drop it into Diablo 2, then drop Diablo (the demon) into Need for Speed, then have my need for speed car show up on another planet and upgrade it into a space ship, while the space ship takes me to fight some mega aliens. All while offering a coherent & unique experience. As you play, the game saves major forks in your story & game genre, so you can invite/share your game recipe with other humans to enjoy.
Also, when are we getting a new Spore game? This game is a sleeping giant waiting to be awaken.
Last year we performed a huge data migration at Capacities. We moved our entire database system from a dedicated graph database to Postgres.
While graph DBs are powerful, we found that for our scale the overhead was massive. By optimizing our data model in Postgres, we reduced database costs by 90% while making our database system more robust and scalable.
In this article, I walk through the process and the challenges we faced.
Going back to the Department of War, and the Secretary of War, is probably a good thing overall. Not to incentivize or motivate war - nothing's changing at all, but to be more forthcoming about the purpose of the department. They're not defending anything and haven't for many decades - geography and nukes take care of that side of the equation, more or less, by themselves.