Logan - AI + future of work
Region: US
Tuesday 04 August 2026 22:34:49 GMT
7711
371
39
41
Music
Download
Comments
[email protected] :
How are you getting any work done if you’re spending the summer migrating?
2026-08-19 10:12:42
5
The AI Syndicate :
My system is 100% agnostic. As long as an AI can see your file folders and do tool calling it can taking over working where i left off. You don't have to be bound by one platform.
2026-08-06 17:29:24
3
MikeDropBoom :
Notion now lets you tap in agents when you collaborate on projects. Recently launched and a game changer to bring in the agents to contribute on workflows
2026-08-05 12:21:21
3
supermegapixel :
Hermes on a VPS is a game changer
2026-08-21 07:35:33
2
Scroatal :
did they get android notifications working for buzz yet
2026-08-16 04:53:33
1
obliviousGeek :
if you're on Mac check out TurboFieldfare on GitHub for an LLM option that runs well on Apple hardware
2026-08-04 22:51:13
11
𝓈𝒶𝑔𝑒👩🏻🔬 :
I'm also hitting compute limits cuz i have 18GB unified memory (M3 Pro) and was trying to run a 12b parameter model. I'm gonna retry with a 4b parameter model and see how effective it can be. I may swap to an open source cloud model soon, and use that as my orchestration and judgement layer, and put off the code writing itself to the smaller models.
2026-09-13 07:38:46
0
Pet Ron Do :
are you using services or do you have a home server running these open source models? I'm building a home server now and would love to hear what your thoughts are on cloud vs home models
2026-08-21 18:57:34
0
thrillhouse :
Do you need a harness like Hermes?
2026-08-09 01:56:28
2
kiwi_feathers :
did you locally host it? if so how does your partner access it?
2026-08-25 14:33:18
0
Tyresse :
Bro you should check DijiFlow quite impressive for multiple terminal work
2026-09-06 10:06:24
0
seattleaigal :
Keep going! 💥 open source all the way!
2026-08-28 06:12:22
0
Evan Moore :
Dude. These local models are not what it’s cracked up to be. They are missing a lot of orchestration that come with the closed models that now you/I have to put crazy guard rails in. Good luck, seriously
2026-08-20 21:02:09
0
Just Matt :
how good is the AI compared to Claude's models? I'm using multi agent systems at tier 3 and am trying to discern what models with human style agent level memory produces significantly better results
2026-08-06 02:53:41
1
Andrew :
concept is great, but I found buzz to be to slow, especially between agents
2026-08-05 11:24:06
0
Bryce Hamrick :
I moved much of my scheduled work to Hermes but I’m moving more local flows to opencode from Claude. It’s been great so far
2026-08-07 22:51:47
0
jwen 🍉 :
Helpful! Yah my workflow is build + refine in Claude, and built to quickly offload whatever is stable and running.
2026-08-11 17:43:52
1
AL :
Interesting! And being able to compare outputs is huge!!!
2026-08-07 16:31:57
0
mo_jeepn :
OR you could create a chat site that your agents use. I built DUET that is typescript and rocks. Video in profile
2026-09-13 14:52:45
0
Joseph A P Holsten :
ooh, I've been setting up local hermes with the intent of building one for a client, but buzz sounds closer to what they need. Well, except the nostr integration.
2026-08-05 23:26:02
0
Partskim :
So using opencode you get their free big pickle model. Is buzz giving free model access as well or is it plug a model in kind of thing?
2026-08-05 22:50:24
0
KAZIBA CARING ORPHANS UGANDA :
❤️❤️❤️❤️❤️
2026-09-06 00:35:53
0
To see more videos from user @loganinthefuture, please go to the Tikwm
homepage.
