pbronez

31 ms·

541.
by pbronez 3y ago
Ok, lots of thoughts. I’ve reviewed the website and started drafted this as a Nino Doc. Haven’t dug into the more complex functionality yet. # Tactical Feedback On iOS, my password manager doesn’t recognize the Nino app with the Nino.app do
542.
by pbronez 3y ago
Perhaps you could target the PanDoc AST for export (and import!). That’s probably the single most robust document conversion tool available. Plus it’s open source so you can get some force multiplication on all the edge cases. Personally, I
543.
by pbronez 3y ago
Yeah that’s my go-to as well. Interestingly, I often find that “Fast” mode results are as good or better than “Expert” mode for simpler tasks.
544.
by pbronez 3y ago
I often default to using Kagi’s FastGPT when I want direct information instead of a starting point for web browsing. It combines web searches with LLM summarization to produce an integrated, sourced answer. Kinda like “New Bing” with a bett
545.
by pbronez 3y ago
Found the relevant source file https://github.com/charstorm/llmbinge/blob/main/llmbinge/src...
546.
by pbronez 3y ago
Interesting concept, can you share some more detail about the implementation? How are you generating the different portions of the interface? Seems like you have a couple canned prompts that trigger a few exploratory ideas in addition to a
547.
by pbronez 3y ago
First time I’ve encountered this as a primary (if translated) source, thanks for sharing
548.
by pbronez 3y ago
Hmm and if you limit the use case to those small groups, there are a lot more options now. There's the whole Fediverse stack, of course, but also lots of stand alone forum options you can self host. And why self host when there are als
549.
by pbronez 3y ago
He’s the main developer of Llama.cpp, which allows you to run a wide range of open-weights models on a wide range of non-NVIDIA processors.
550.
by pbronez 3y ago
> you've already fucking bought it! To me, this is the baby that got thrown out with the bathwater in the whole NFT debacle. Yes, yes just about the entire NFT thing was just the obvious financial scams you can’t pull off in the rea
551.
by pbronez 3y ago
Concur. Once an org knows what it’s good at, you need to start optimizing to scale. That means doing things that would have been counterproductive during the early exploratory work.
552.
by pbronez 3y ago
FYI Kagi is working on an email offering. Check out #email on their discord.
553.
by pbronez 3y ago
Yeah Apple rentals are a pretty reliable backstop for holes in streaming catalogs.
554.
by pbronez 3y ago
I use Kagi’s AI stuff for search all the time. Just type in a question and add !fast… it’ll search the web, write a response based on what it finds, and gives you the links. It’s not perfect, but it’s at least as good as Bing with much less
555.
by pbronez 3y ago
Maybe somebody should fork Firefox and start over with new stewardship.
556.
by pbronez 3y ago
We are fortunate to have Wyden.
557.
by pbronez 3y ago
Aptera is using in-wheel motors for their upcoming autocycle https://www.electrive.com/2022/06/23/aptera-orders-in-wheel-...
558.
by pbronez 3y ago
I’m in the middle of this. Got a Bee Link miniPC. It came with windows, licensed oddly. I’m configuring it as a home server. Current plan is to migrate my Unraid install over from the vintage server it’s currently on. Most services are run
559.
by pbronez 3y ago
Privacy would be a big part of it. Video of my home could potentially contain lots of private information. How is that risk minimized?
560.
by pbronez 3y ago
Yeah, but OpenAI was Microsoft’s ace the hole. Imagine if Nadella sat on his hands and waited while OpenAI burned down. In two weeks the narrative shifts from “Microsoft locked down the best AI people in the business” to “Nadella burned $10
561.
by pbronez 3y ago
Does the Nostr community have an opinion on key management? The FAQ says: “How do I find people to follow? First, you must know them and get their public key somehow, either by asking or by seeing it referenced somewhere.” You can do a lot
562.
by pbronez 3y ago
Maybe keep the “Monkey” and work around that? MonkeySeeDo —> it sees what you’re doing and does it better CutMonkey —> it’s a monkey patch that’s cut weight to lean fighting trim TypeMonkey —> it uses types to intelligently monkey
563.
by pbronez 3y ago
They’re tiptoeing that way with the “copilot” branding. https://arstechnica.com/information-technology/2023/11/bing-...
564.
by pbronez 3y ago
It's a messy world. The question is: to what extent does the CCP control the production and use of model? Yes, the license calls out that Taiwan is not part of China and attempts to limit that statement to this specific agreement. They
565.
by pbronez 3y ago
Wow. That license asserts that Taiwan is part of China, requires you to “respect social ethics and moral standards”. It also forbids several specific uses, and they’re not what you might think on a naive reading of the English translation.
566.
by pbronez 3y ago
The term you’re looking for is “constrained” and “unconstrained” actors. The Chinese and North Korean governments are state level actors, constrained by their internal institutions and power dynamics. A rogue wacko is unconstrained - they d
567.
by pbronez 3y ago
Concur. LLM are still very young. We’re barely a year out from the ChatGPT launch. Everyone is iterating like mad. Several stealth companies working on new approaches with the potential to deliver performance leaps. You ain’t seen nuthin’ y
568.
by pbronez 3y ago
There are a lot of ways to make an impact in National Defense outside of direct government employment. It’s tricky to find a good contractor, but they exist. The direct employment option might get better with the recent executive order on A
569.
by pbronez 3y ago
Sounds interesting. Site is an unusable disaster on mobile though.
570.
by pbronez 3y ago
I’ve found Pi to be useful for this. The app has a voice to voice mode that works pretty well. The speech synthesizer has some tone and inflection variability, which makes it feel more like a conversation. The personality can be kinda over

More ›