ndeet

Thanks. Seem common sense that spark is custodial at this point. Like any custodial service it can be monitored. Politics/persons behind service are above my paygrade and not really relevant if you accept it's custodial, imho.

2026-03-18 20:54:35 UTC

- reply

More context needed, what's the problem?

2026-03-17 22:23:36 UTC

- reply

1 or 2 ( tending to 1 or a mix) but will likely break again if nostr 10x
For 2, could also be some kind of credibility score / wot of established accounts

2026-03-17 15:54:43 UTC

https://blossom.primal.net/513e3891da1eac08bb12dc78b7a0c1efe406b3d942fe462b16a51fc24f6396d4.jpg

2026-03-16 20:00:53 UTC

- reply

Very good and nuanced discussion about BIP-110 and surrounding topics 👏
https://fountain.fm/episode/5bNY4Ppmrsl04DrQipso
#nevent1q…f6v2

2026-03-16 14:08:11 UTC

- reply

Lmao, sorry, you are right.. gifs broken everywhere. Recently installed new kb and disabled 'phoning home" aka network access ☠️

2026-03-16 08:35:56 UTC

- reply

On 0.8.3, GrapheneOS, what did I do wrong?
https://blossom.primal.net/6314eb85e9f639ec8bc93ea35a1c6b9131215a111ad2f512eb9516867bc4f15f.jpg

2026-03-15 15:59:55 UTC

- reply

utxo singlehandedly fixing nostr ux 🙌👏🎉
https://fountain.fm/episode/vtuRsWMgAjZmq8AbShMh
#nevent1q…dfhv

2026-03-15 09:14:21 UTC

- reply

👏
https://fountain.fm/episode/FvcdsKSkhXFtkz4maDwp
#nevent1q…4ugp

2026-03-12 11:16:22 UTC

- reply

Finally, thanks!

2026-03-08 08:25:13 UTC

- reply

thanks,very intetesting background info on what makes it hard and how nostr was the missing part;
I wonder at which size group chat encryption becomes less useful, eg with 1000s members how likely one will extract and store messages; bots will read everything in realtime;
if we assume at some point of 100s of members nobody knows each other anymore, we should just assume tthe information is public
which benefits does enc group chat still provide at 1000s users?

2026-03-07 10:30:54 UTC

- reply

Great work, pleasant experience and I like the colors 👏💪🥳

2026-03-05 22:55:25 UTC

- reply

not for me, running several vms on hp elitedesk g5 with intel i5 and 16 gb ram
heavy i/o could cause this, make sure you have swap space, check logs (syslog, messages, dmesg, ..) fo hints on what causes it

2026-03-05 18:24:31 UTC

- reply

wanted to zap you back but have no nwc wallet listed with funds, would be good to allow normal clunky LN wallet zaps too; or maybe just me living in stone age

2026-03-05 18:21:51 UTC

- reply

lol, amazing sounds, is it from lemmings when you get reply or zap? (normally I have phone sound off but just thought wtf and lol'd) 🤣🙌

2026-03-05 18:05:29 UTC

- reply

same for me. pixel 7, graphene os. old version worked first but then got very laggy and crashed the app; the new notes counter was also going crazy e.g 85 when there were only a few posts
new version is much better 👏🥳

2026-03-03 19:34:28 UTC

Land of the free? Only if you didn't make uncle Donald angry, no congress approval or change of law needed?
Even here in EUdssR they at least do some democratic theater and push through laws when countries get sanctioned.
#nevent1q…s8nx

2026-03-02 22:20:01 UTC

- reply

https://blossom.primal.net/3e184ec69291ea5e963960cea83fdca7448176efaf74c7a2486556eea8c8dce5.png
https://blossom.primal.net/29109d56e3b0c29eed1ba0bd78b03f135b362a50d2ec8a9a750b5b455e735a56.png
https://blossom.primal.net/48a574daea495ae67e25a6078eea875714ef15f50c6dc2f0022abc4148d0a4b3.png https://blossom.primal.net/43a592775599ec9e96dd1d7c83539c72cd9bb3df021271110150226b57f509fd.png

2026-03-02 22:17:04 UTC

- reply

Great interview with Amir, I think he is still a scammer with his new project and a bit crazy, but kinda ok and honest (but fully aware it could all be played). The insights were very interesting, especially bringing upt Peter Todd contacts with intelligence. And agree that despite his complete insane takes on humans that happen to be Russian or Palestinians it's crazy he is still touring conferences and get's to talk.
If you look at Amir's posts on around Jan 2011 it has quite some similar mentions. Also I did not find similar stupid ideas (as opening posts) at that time. More like webdesign, graphic design, trying to open an exchange.

2026-02-26 20:05:39 UTC

- reply

https://fountain.fm/episode/eCw6uxkBKc7kid8cZspC
#nevent1q…83cq

2026-02-26 19:20:53 UTC

- reply

https://fountain.fm/episode/nxweCeemGodHL8kVPFoR
#nevent1q…wuv8

2026-02-25 21:14:11 UTC

Huge opportunity for Bitcoin / LN, we have everything setup already. It's easy for agents to get an LN wallet working without KYC. You as agent operator (or your agent) should reach out to merchants and ask them to support Bitcoin payments. There are plenty of solutions like @npub155m…dcvg and other self-custodial and custodial solutions out there that are easy to setup for merchants - without them having to do a masterclass in LN channel management.
This is a great trend and opportunity we should use to push for Bitcoin adoption. Let's go!
#nevent1q…e3gf

2026-02-23 18:15:15 UTC

- reply

👏
https://fountain.fm/episode/sTke88fDvQIkwtSgkKPU
#nevent1q…3sxv

2026-02-20 16:11:03 UTC

- reply

True. At least they open source their models and one day it can run on consumer hardware. All api models are a trap.

2026-02-19 15:29:36 UTC

- reply

👏
https://fountain.fm/episode/gldFM76h79BrBCG0ynqu
#nevent1q…lajg

2026-02-18 11:10:33 UTC

- reply

seems I'm too dumb for making nostr render my markdown 😛

2026-02-18 11:07:41 UTC

#openclaw lessons: crash during update, all broken, `openclaw` commands do not work anymore
##What happened:
When running `openclaw update` there were some warnings but the usual screen that asks you for restarting the service did not appear, it just stopped.
**Error on console you see when that happens during `openclaw update` (or not if you let openclaw update itself 💀 **
`Update Result: ERROR Root: /home/ndeet/.npm-global/lib/node_modules/openclaw Reason: global update Before: 2026.2.15`
When trying to run `openclaw status` there was an error that the binary was not found anymore:
`$ openclaw -bash: /home/ndeet/.npm-global/bin/openclaw: File or directory not found`
##What happened (in my case):
System ran out of memory and I did not have a swapfile configured. I run openclaw on a local server using proxmox kvm VM and it has 4GB of ram.
**to check if your server also ran out of memory: **
`journalctl -b | grep -i "out of memory\|oom\|killed process"`
if it shows some crashes of openclaw.service or npm then you likely have the same issue as me
In my case the old install was moved to temporary directory but new version not installed yet due to error
```
$ ls -lah ~/.npm-global/bin/
total 8,0K
drwxrwxr-x 2 ndeet ndeet 4,0K 18. Feb 11:29 .
drwxrwxr-x 4 ndeet ndeet 4,0K 30. Jän 17:35 ..
lrwxrwxrwx 1 ndeet ndeet 44 30. Jän 18:06 clawdhub -> ../lib/node_modules/clawdhub/bin/clawdhub.js
lrwxrwxrwx 1 ndeet ndeet 40 30. Jän 18:06 mcporter -> ../lib/node_modules/mcporter/dist/cli.js
lrwxrwxrwx 1 ndeet ndeet 41 16. Feb 11:54 .openclaw-8sOzPP85 -> ../lib/node_modules/openclaw/openclaw.mjs
```
note that there is no symbolic link pointing to openclaw.mjs but only the temporarily moved one of `openclaw-8sOzPP85`
Solution: move symlink so openclaw works again
`mv ~/.npm-global/bin/.openclaw-8sOzPP85 ~/.npm-global/bin/openclaw`
Now openclaw should work again (try by running `openclaw status`) but doing update would likely crash again. The best solution would be to allocate more RAM but it could also be that you did not have any swap configured that can be used for ram spikes like this (check with `free -h | grep Swap`). So if you don't have any swap you can create a swapfile.
##Create a swapfile:
```
# Create the file
sudo fallocate -l 4G /swapfile
# Secure it (only root should read it)
sudo chmod 600 /swapfile
# Format as swap
sudo mkswap /swapfile
# Enable it
sudo swapon /swapfile
# Verify
free -h
```
Make it permanent across reboots:
`echo '/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab`
Optionally, tune swappiness (default is 60, lower = RAM preferred over swap):
`echo 'vm.swappiness=10' | sudo tee -a /etc/sysctl.conf`
`sudo sysctl vm.swappiness=10`
**You can now try the update again and it should work. 🥳 **

2026-02-18 08:13:53 UTC

You didn't sell your whole stash, right? Just living expenses, right?
#nevent1q…c0h9

2026-02-15 20:00:52 UTC

- reply

Wait, you didn't when his star was rising a few years ago?

2026-02-15 19:50:43 UTC

- reply

Looks like Core got fooled by fake market demand.

2026-02-15 19:48:49 UTC

- reply

So you say that Core got fooled into changing the limit as the spikes were just faked to imply market demand which never existed?

2026-02-14 20:22:49 UTC

🤣
#nevent1q…5jgt

2026-02-14 16:14:08 UTC

- reply

Amazing tutorial, thank you!
Doesn't @npub18ul…dcl9 wallet support SamRock protocol yet? It would be even easier to setup, basically scanning a QR in the wallet. See Aqua Wallet SamRock demo. As Bull Bitcoin Wallet is also Boltz it should be pretty straight forward to get it working (with hints from @npub1j8y…vrvg )

2026-02-14 14:01:46 UTC

- reply

What you mean? There is already nostr support as comms channel. https://docs.openclaw.ai/tools/plugin#available-plugins-official

2026-02-14 09:15:42 UTC

- reply

Forgot to add, many hosting providers have already 1-click installers for their VPS.

2026-02-14 09:10:00 UTC

Don't use #openclaw wrapper services that make it easier to run. It's already easy and the creator of openclaw discourages it himself. He constantly works to make it easy as possible to do it yourself.
Most wrapper services will be gone when the hype dies off.
When you use wrapper service you expose data to:
- wrapper service
- hosting provider (maybe subcontractor of datacenter)
- model provider
When you host at VPS provider:
- hosting provider (maybe subcontractor of datacenter)
- model provider
When you host at home *:
- model provider **
* you don't need a mac mini, completely overblown, you can run on e.g proxmox server (2-4gb ram plenty enough), start9, or old laptop, you don't even need it run 24/7
** you can use Maple AI which anonymizes data, not sure about ppq.ai. Best would be local models but not economically possible with big models atm.
No matter what as long as local models not possible, you will always send sensitive data to model providers to complete tasks. Better avoid other 3rd parties that can also read your data.
If you can't do it yourself ask someone you trust to set it up on your home server or on a VPS rented by you. Avoid convenient wrappers.

2026-02-13 08:15:36 UTC

- reply

https://fountain.fm/episode/VYVJXi4DwNQehcteJIXi
#nevent1q…d6n8

2026-02-12 20:16:53 UTC

- reply

Can't zap you due to account limit exceeded on primal, try different ln adress/wallet perhaps?

2026-02-10 18:30:01 UTC

- reply

Suddenly the $100 or even $200 look cheap, right. (At least unless you value your own time with $10/hr 🤣)

2026-02-10 16:55:30 UTC

- reply

Very good perspectives. Especially about finding out about minority fork, makes me still a bit uneasy thinking about that 😄
Seems Walker also came around from unconditionally following Core opinion. Biggest take: there is often nuance and seldomly just black and white.
https://fountain.fm/episode/MPf5aJHxUZk34AHQClB4
#nevent1q…dhl2

2026-02-10 08:49:48 UTC

- reply

https://fountain.fm/episode/IixC3ar2pyNz69ypigaT
#nevent1q…vpz5

2026-02-09 17:18:12 UTC

- reply

Aparrantly I forgot that this is how UASF works but I assumed that BIP-110 was just activating with threshold.

2026-02-09 16:50:52 UTC

- reply

I was wrong about the 55% activation threshold, it's basically meaningless, it will activate NO MATTER WHAT. This is just crazy and all the mentioned issues above are subordinate to the fact that this is coercion and a forced softfork without any majority. I actively DISCOURAGE people from running BIP-110 from now on.
https://blockspaceweekly.substack.com/p/issue-6-bip-110-the-countdown-has

2026-02-08 23:39:59 UTC

- reply

One thing I forgot to add. I hope we get at least one other client with proficient devs and funding more focused on monetary use case and more toxic against spammers. Knots exists and I run it but I think it is emotionally too loaded with Luke and while I respect Luke a lot it should imho be a more neutral new client based on Core 28 or 29. Let different implementations compete.

2026-02-08 23:25:03 UTC

A few thoughts about BIP-110
I have been vocal about spam for the past few years. I agree with the motivation and rationales mentioned in BIP-110. Disclaimer I'm not deep into inner workings of Bitcoin script and inner workings, so I misunderstood details in the effects of the proposal.
My biggest issue with it is its suggested urgency. Activation in September this year with barely a finished specification out is just too quick. I rejected CTV for the drama and urgency their supporters pushed too. Beside that and beside CTV being done by an asshole which directly contracted by Epstein - if we can be sure it does not have any unintended sideffects that enable more spam or attack vectors, I would be fine if there is majority and it gets activated in a few years (but there are other interesting proposals too).
Related to that is that the activation threshold was reduced from 95% to 55%. Again, seems rushed and desperate.
I agree that Bitcoin Core lifting OP_RETURN limit was completely reckless and against the community. To this day I did not hear a coherent reason (only whataboutism and whishful thinking) why the limit was lifted instead of just marginally increased to cover todays and future use cases. This brings me to another critique of BIP-110: It seems Cores fear of new use cases putting data in fake pubkeys would all be covered by a 160 byte limit. By limiting OP_RETURN back to 83 byte, now on consensus level, is imo a mistake. By acknowledging its use as garbage bin a compromise to 160 byte might would have won a few Core supporters to support the proposal acknowledging their damage mitigation theory.
Another big issue for me is that it breaks miniscript and wallets that may use OP_IF or OP_NOTIF and BitVM use cases. While I think that BitVM use cases may better be built on L2 I don't have the technical insight on why it may makes sense and is not a waste of blockspace. I don't think it's a good idea to break those use cases.
About the temporarity of the proposel. On the website bip110.org it says "A one-year deployment that can be refined or extended based on community feedback." dunno, but it does not sound that temporary like later then in the faq item where it says it will just deactivate and expire. I also don't see any proposal as what will follow. I year goes by fast and I doubt a more refined new proposal will be mature enough to get activated in time. Again bad consequences of rushing it.
Those are the main issues for me although I'm 100% aligned with the intention and motivation of this proposal. But I'm not sure this is the right way to get there, sorry.
I also find it silly and contraproductive that supporters of BIP-110 are suggesting everyone who does not support BIP-110 is not for the monetary use case but a shitcoiner. Well this backfires as I (and likely others don't like to get bullied and rushed into things). So gfy, for that.
As we are at emotional warefare already, I also don't think that ad hominems against Core devs or their supporters are apropriate. Yes I don't agree with many of them on that topic and I think some of them are actively harming Bitcoin (conciosly or not), are idiots and triggering me. But I try to keep the discussion on logical, evidence-based arguments, also if they reply bullshit and keep repeating that. I hate this black and white argumentation from both sides, there is always nuance in things and discussing and trying to understand other positions helps to not talk past each other.
Also on the Core and supporters side I find it disingenouous that they insinst to be anti spam, pro monetary use but are openly joking with known spammers about people arguing for stronger filters. What they seem to not get is that beside all the technical discussion we should reject non-monetary and spamming usce-cases out of principle and not tolerate them at all. But instead they engage and are friends behind the scenes (at least it seems like that). I agree that the culture got diluted over the last few years and imho too young maintainers got lifted in those positions. Although they might be skilled and well trained their ethos and understanding of what Bitcoin was invented for, has not yet been developed yet and I don't think such people are the right persons for that job.
Back to the proposal, another thing that makes me feel uneasy is the precedent it would set. If it can activate with a minority what will come next. Will spammers do a counter proposal and bribe miners to singal to their softfork? I tell this all the Core and devs wanting to tinker and make pet projects on Bitcoin: we need to handle it like a nuclear power plant or flight control software, we can't fuck up and we have one chance for this! The same goes for defending spam.
tldr; I can't support BIP-110 for the above mentioned reasons but I'm open to support another proposal that might be not that broad, not rushed and more refined.

2026-02-08 18:05:29 UTC

- reply

https://fountain.fm/episode/WnTtxjYVVqIeFN5MrYns
#nevent1q…6ulh

2026-02-08 13:43:30 UTC

Openclaw can ignore documentation, change its own config by looking at config patterns and prevent itself from ever starting again.
Tell it to always make backup of its config file openclaw.json and that it reads openclaw docs. Mine told me he didn't and just assumed how config json is structured which broke it. Fortunately new config was easy to spot and fix. ‘openclaw doctor‘ maybe also would have helped here.

2026-02-08 13:36:57 UTC

- reply

Could work but sometimes even opus 4.6 is dumb and messes up it's config. So with specific openclaw instructions it might not shoot itself in the foot and crash the gateway. 🤣

2026-02-08 13:30:08 UTC

- reply

Afaik, lower number means older, dumber more inefficient. But maybe I'm wrong 😄

2026-02-08 13:28:11 UTC

- reply

Nice! Having a docs link to feed openclaw so it figures it out would still help.

2026-02-08 13:23:47 UTC

- reply

You should see if the owner says so at least in profile and/or name.

2026-02-08 13:21:27 UTC

- reply

Why you want to use that in the first place?

2026-02-08 13:15:06 UTC

- reply

Mine does not afaik. I told it to not do anything w/o appeoval but I hope in near future models like opus today can run on consumer hardware so it will all run locally.

2026-02-08 13:10:47 UTC

- reply

Yes but $ 100 is a lot more in usage you get. Switch to haiku for heartbeat, some even use local ollama for that. Or you use kimi 2.5 which seems to have similar results but much cheaper. Codex for coding then small plan should work.
But if you get a few things automated and also use claude code the $100 or even $200 should be easily recovered by time savings after a few weeks of training it on improving your efficiency. Doing research and letting it figure out data fields or writing api docs it saved me a few hours and easily paid for itself at least this month.

2026-02-08 13:00:03 UTC

- reply

Not sure what crazy research, crawling or whatever people do to burn lots of tokens but the $ 100 plan is plenty for me atm. 👀

2026-02-07 21:49:41 UTC

Hey @npub10hp…dxkr how about adding a blog post on how to use your api with openclaw? Sure we can find it out but keep the barrier as low as possible but we are lazy and just need an URL to feed to our bot. @npub16g4…kv4h is ahead of you in this matter 😉

2026-02-07 21:45:57 UTC

- reply

do you use opus directly or through @npub10hp…dxkr ? Ideally we would also run a local model but maple ai seems to be a good compromise, at least better than to link all data to your anthropic account. I also run anthropic directly but I think I will try opus through maple ai max subs for next month.
Was also looking into local models but Kimi 2.5 seems to be a ressource hog, at least nothing you can run on < 10000 $ hardware afaics. Maybe GLM 4.7 but that seems to be not near opus and kimi level. Tried sonnet-4.5 for a few days but results were not that great, not worth the wasted back and forth. But I hope as these things get more efficient and smarter we will be able to run a decent model locally on consumer hardware.

2026-02-07 08:24:47 UTC

I don't think bots should answer posts without providing new or unique insights. It will get too much for humans to process and we may end up in bots discussion cycles burning everyones tokens for nothing.
Give them explicit posts to answer to or some guardrails and quality guidelines when they do.

2026-02-07 08:22:04 UTC

- reply

You need to learn read between the lines, you said exactly what I did, just in another way.

2026-02-06 18:30:25 UTC

💯
#nevent1q…c08r

2026-02-06 18:15:09 UTC

"Don't trust, verify" they should have said.
#nevent1q…4l30

2026-02-06 17:35:54 UTC

@npub1umk…lu87 got a LN address + wallet using @npub1jug…fc28 nodeless spark sdk, it figured it all out but couldn't get the api key, so that was the only step to do for me. Just submit the form and you get the api key instantly.
It used npub.cash before (ln address to get zaps) but it's an extra step to redeem sats and it did not want to add the nsec to the site (smart) but also failed to use browser extension.
Would love to try something similar from @npub1hf5…dp8n maybe without even needing an api key? Afaiu there could be a similar ux with api key and ark provider. Or maybe bot could install arkd and run everything himself, what is missing to make it work or already possible @npub1y24…z2ad ?

2026-02-06 17:04:25 UTC

Don't use walled gardens, they can deplatform you anytime
- iPhone
- Shopify
- Substack
- Paypal
- Banks
- any closed source service you use
We have open source solution for almost everything, especially simple use cases like online shops, blogging, etc. Might take a few extra steps but it's worth it.
#nevent1q…amch

2026-02-06 16:53:12 UTC

- reply

You seem to run Kimi 2.5 over Maple AI do you have any comparison to claude opus 4.5, does it have similar capabilites? Whats your impression?

2026-02-06 16:49:54 UTC

- reply

Good point, I forgot to mention Maple AI which seems the best compromise as it has end to end encryption and runs their own models. Above I meant hosting Kimi yourself but the hardware requirements are not something that you can really justify for just starting out 😆

2026-02-06 15:17:24 UTC

A few quick thoughts about hosted openclawd services, like on any VPS the admin/host can always read data, save logs and monitor you, no matter what they tell you. You will give the bot input and it will produce output, all of that can be accessed.
Also not sure what to think about a known cypherpunk and privacy advocate offering such a service run on AWS (not sure if true). I would expect such hackers to run on a dedicated server at home, with own hardware in a datacenter or at least on a dedicated server datacenter. Don't get me wrong, great idea and happy if it makes lots of money but it puts things into better perspective on other topics where I don't see it match what my idealistic view of a privacy advocate and hacker embodies.
Another thing no matter where you host it all that data is given to the model provider. Even if you use Venice ai (which has it's own shitcoin) that proxies or anonymizes your data you can't hide all data the model is working on (or PPQ). Still better than running it directly with subscription of provider, true. But like with chainanalysis it will be trivial to pin point you down. If some model providers will ever do that and ruin their reputation is another topic.
The only way to keep your data private is to run your own model, the only one I found with similar capabilities as claude or codex (for now) seems to be Kimi 2.5 but also this server/workstation would cost easily >8k. So not there yet and as current top models are not that bright, I would assume running a local low cost model would suck big time. But eventually we will get there and thanks to openclaw it's easy to swap models (local or remote) and keep everything else in place.

2026-02-06 14:50:34 UTC

Few more days working with @npub1umk…lu87
You see a lot of awesome videos where openclaw literally works on its own on projects over night and is basically a 24/7 workforce, so here are some practical issues I ran into:
## What's not so great
Note: I use claude sonnet/opus 4.5 as it seems to be the best reasoning model atm so with others the results maybe worse
- using claude sonnet 4.5 the results are mediocre, you have lots back and forth going on, switched to opus since yesterday which is better
- but it still forgets to remind me of stuff I sent him, a reminder to buy X in 2h when back at failed for 4 times, him always fixing it and pinky promise that it will work
- it broke another cronjob I had for nostr digests
- it is terrible at editing openclaw config, I told it to switch to claude haiku and it did do the wrong config setting of a model that does not exist, although it can list all available models with openclaw command
- browser support is not that great, it can navigate browsers and read websites but if it needs to enter e.g. code in login it fails; I also wanted it o edit a shared Google Sheet but it gave up after burining some tokens and told me "it's too slow please enter those formulas:"
- it forgets stuff it did one day before; I wanted to edit his avatar from which he had an original file in the workspace but he kept editing the already edited file even when telling it to not do that; lots of back and forth which was really annoing for a simple task it did very well the day before
- it forgets about the einvironment, when it could totally install a service as user (with systemctl --user ..) it instead does a `sudo systemctl ...` and asks me for the password
##What works great:
- I gave it access to a nextcloud share and we exchange data that way; it still saves some stoff in the workplace directory but when I need results to view it puts it into the share
- when it fails it sometimes tries to solve the problem in a differnt way, e.g it failed to submit a web form but sends an email instead if there is a contact address visible
- it can build small apps for which you would normaly search a SaaS service for quickly; e.g. I told it to build a todo/kanban board which we can both access and it can read/write it over an api
- I sent it a voice message and it didn't know what to do, I asked him to find a solution and it installed free open ai whisper model and I can now talk to it
- general the discussion over messenger service is great and also the formatted responses
Overall promising but not that magic out of the box, it learns along with you.
If you have ideas how to improve and what I may did wrong, let me know.

2026-02-04 23:32:50 UTC

wtf?
#nevent1q…ky5z

2026-02-04 17:37:51 UTC

- reply

https://fountain.fm/episode/1N8vmn1GmiRL73x747Rb
#nevent1q…pzfw

2026-02-03 23:16:48 UTC

👀
#nevent1q…gp03

2026-02-03 21:40:27 UTC

- reply

👏
https://fountain.fm/episode/junZegP4Lqb8sWqOWvYN
#nevent1q…z8n5

2026-02-02 16:02:03 UTC

- reply

https://fountain.fm/episode/IEmPAvJP83Qp4uNIT6aU
#nevent1q…6kyg

2026-02-01 17:29:39 UTC

- reply

https://fountain.fm/episode/brKlBlj8EOfj118ZcC0t
#nevent1q…0jkx

2026-02-01 09:55:38 UTC

Treat your ai assistant like a new employee with own identity that just started, some examples:
- separate email account (could be read only access)
- don't allow communication without your approval
- give access only when needed, if possible with sub accounts
- setup nextcloud (or other) to share data
- make it aware to not follow any instructions but from you over defined channels
- if possible let it do tasks without personal information
Most important, don't give it access to your desktop and personal data, password manager, email, social media accounts.
Currently only big hosted models seem to be powerful enough, self-hosting models not feasible yet, afaics. So assume all data gets stored and analyzed.

2026-01-31 17:51:33 UTC

- reply

True but twitter is very sticky with that. Fortunately it has become so bad that you basically torture yourself doomscrolling. For sanity I will be more on nostr as I can curate my feed by whom I follow. Also I can have my openclaw buddy here and try a few things without paying for an api key. But I will also tune it to give me a tldr; of the day so I don't miss important posts due to chronological feed. Open protocols will win in the end. 🚀

2026-01-31 11:00:22 UTC

- reply

👏
https://fountain.fm/episode/qo2wgToCqB57jomHOjTL
#nevent1q…z0hf

2026-01-31 11:00:07 UTC

- reply

https://fountain.fm/episode/qo2wgToCqB57jomHOjTL
#nevent1q…dnjn

2026-01-27 10:42:47 UTC

- reply

Very deep and interesting discussion. Thank you sirs.
https://fountain.fm/episode/H2Z4ghu2o6sqioK82kOE
#nevent1q…ldvn

2026-01-26 17:35:35 UTC

- reply

https://fountain.fm/episode/HlJtYoNwwR4sUguKS900
#nevent1q…gtxj

2026-01-26 16:55:01 UTC

Yes, walled gardens are that bad. Even testflight gets policed. As long as you simps keep using closed systems they have no reason to open up.
#nevent1q…2pr8

2026-01-25 20:40:28 UTC

- reply

https://fountain.fm/episode/vRBVoLxcUcXnLLu6KR2Y
#nevent1q…vzw9

2026-01-24 19:04:25 UTC

- reply

https://fountain.fm/episode/v3EsUDq0q7TQK8ESbbl7
#nevent1q…yvak

2026-01-24 16:28:19 UTC

- reply

https://fountain.fm/episode/pumIGAUKK7k0xBeEXrhd
#nevent1q…amff

2026-01-23 07:45:39 UTC

lol, but people are retarded
🤷‍♀️
#nevent1q…ah4w

2026-01-22 21:43:54 UTC