verdverm

17 ms·

511.
by verdverm 2mo ago
you dont need open claw, any regular harness will do these days the difference is mainly that you can interact with the agents via a phone chat app like telegram Scion is an alternative I have my eye on
512.
by verdverm 2mo ago
Does the point need to be a making money? I know some people who use intercommunicating agents (claw style) to automate boring personal life stuff like staying on top of their bills or researching a vacation
513.
by verdverm 2mo ago
title is editorialized, it should match the article doesn't have to be an either or, the buffet of options is getting longer and cheaper, token generation is commodity stuff
514.
by verdverm 2mo ago
Isn't this how we ended up with algos that advantage anxiety and doom scrolling?
515.
by verdverm 2mo ago
search as we knew it is over, better to have your own agent handle this process than google's
516.
by verdverm 2mo ago
you dont know me, so I don't know on what basis you make this claim, boycotting Elon and his companies is a real thing people do for ethical reasons, see Tesla sales as a data point
517.
by verdverm 2mo ago
It's typical to use a special STT model (audio in only), which will not be able to clean up afterwards. If you are using an LLM for the STT part, you're leaving stuff on the table
518.
by verdverm 2mo ago
its not a meaningful step, its a bare minimum (ACL) token acceptence of reality, I don't get the sense they actually believe in private data on social its designed in such a minimal way that the app devs will have to implement on top f
519.
by verdverm 2mo ago
I'm not keen on OpenCode long term after seeing their CEO engage in antics and their discord being rather unprofessional, but it is a great harness for the time being, still has gaps though
520.
by verdverm 2mo ago
second this notion, fresh sessions and markdown are far superior to compaction, which has high likelihood of wrecking the context
521.
by verdverm 2mo ago
In a sane world, one should be able to choose not to do business with a company for any reason there are plenty of alternatives, 100s, so I don't think antitrust is on the table, think of this more like a boycott of the world's ri
522.
by verdverm 2mo ago
fyi, Elon Musk is not popular around here re: DOGE, politics, and the needless deaths on his hands
523.
by verdverm 2mo ago
with wireless tech, you can embody Ai in just about anything, waiting for that hacker post about turning their toaster into a thinking machine, I have a Anki vector I've been meaning to do this with (has camera, speakers, microphone, a
524.
by verdverm 2mo ago
I only use open models now, I really think the era of open models is upon us, big or small, but I also agree small models reached the point where you don't have hand hold them with qwen 3.8 27B
525.
by verdverm 2mo ago
Tragedy happens every day, this is likely one of those freak events we cannot do anything about. Millions of people die from hunger and malnutrician every year. Perhaps we could focus our energies on that were it not for the media (us by pr
526.
by verdverm 2mo ago
have another model do a pass to clean it up, saw a demo of local STT where someone did this, can fix a lot of things, especially with gotchas for the STT model in a clean-transcript.md
527.
by verdverm 2mo ago
one fundamental issue in atproto is the missing permissions layer, the proposal is insufficient for many use cases, graze was shut out of their main financial path by Bluesky (paid creator posts, Bluesky cut off talks) tl;dr - people want c
528.
by verdverm 2mo ago
This is what the DC budget (Washington, as opposed to data center) is for.
529.
by verdverm 2mo ago
It seems really unlikely there will be any winner takes all, I'm not sure long term model vendor will get 10% of the market. I expect token machines to look more like phones and cars, with a lot of user preferences, then the cloud with
530.
by verdverm 2mo ago
I've always wondered why everyone flocks to SV's latest darling company. Have we not learned from our history of glorifying these SV darlings that turn hostile?
531.
by verdverm 2mo ago
per-token, supposedly FirePass v3 is coming, soon (tm)
532.
by verdverm 2mo ago
The data centers are for inference time serving, and as long as the GPUs go better, it shouldn't really matter. Token prices should go down 10-100x over the coming years, and then we will use more, because they have more capacity for t
533.
by verdverm 2mo ago
if you download them, you can remove all internet access for the model going forward on the task, don't want to http proxy like OpenAi now do we?
534.
by verdverm 2mo ago
Trust is a two way street, why do you trust them (Ant) if they do not trust you? Have they done enough shady things yet to break it? Are their models really that far ahead it's worth it?
535.
by verdverm 2mo ago
I wouldn't add it yet, just because Salesforce wants us to doesn't mean we will.
536.
by verdverm 2mo ago
Have you tried others? You're leaving a lot on the table using Rovo, it's just not nearly as capable
537.
by verdverm 2mo ago
I never understood why GH didn't introduce chat before acquisition, I wouldn't use it post tho (b/c teams)
538.
by verdverm 2mo ago
no trying required
539.
by verdverm 2mo ago
actual video, he goes off on a number of things https://www.youtube.com/watch?v=7NtOYzDJt9U I saw Vader sing "pedophile in the white house" earlier this year, it was pretty good too
540.
by verdverm 2mo ago
OpenCode is highly unprofessional, especially when compared to alternatives like Fireworks. I still use the harness and pay for their cheap af plan tho, but have no loyalty, would really like some alternatives.

More ›