Slazerslaze

@Techaktien1 @Jako_Leee Passend dazu => youtube.com/watch?v=C2rtm3…

@DavidOndrej1 It is indeed a very ugly design. I'm getting some Jaguar vibes here ;-).

ugly as shit

@sparbuchfeinde Die Frage ist, ob sich dann irgendwas ändert?

Wir haben Ende Mai. Schon im September könnte unsere Regierung vor dem Aus stehen. Durchaus wahrscheinliches Szenario: > Erster AfD Ministerpräsident mit absoluter Mehrheit in Sachsen-Anhalt > SPD fliegt in Sachsen-Anhalt komplett raus > SPD verliert Ministerpräsidenten-Posten in Mecklenburg-Vorpommern Die letzten beiden Punkt kann Lars Klingbeil politisch nicht überleben. Die Union kann sich ggf. noch irgendwie durchwursteln. Aber auf die SPD kommt ein Erdbeben zu. Nicht unwahrscheinlich, dass nach Klingbeils Rücktritt die komplette Koalition zerbricht.

@DaveShapi hopefully, humanity will go through this and be stronger and a bit more intelligent then the actual human OS => feels like we are in the Windows Vista implementation at the moment...

Humanity is about to go through a major, collective positive disintegration. It's gonna be messy.

@CWRoehl Ich finde die Dame ist massiv überbewertet => entsprechend sind auch viele (natürlich nicht alle) Ihrer userungen Fehlallokationen ;-)

Investieren in Indices = Fehlallokation von Kapital? ...sagt ausgerechnet Cathy Wood, deren $ARKK Innovation #ETF seit Start im Herbst 2014 nicht einmal den gleichgewichteten Nasdaq 100 ETF geschlagen hat.

@SimonHoiberg That’s a nice solution! Thanks for the tip!

One tiny n8n node makes AI agents way less scary: Wait. Most people think of n8n as the thing that runs workflows. I use it as a proxy layer around OpenClaw. OpenClaw can decide what needs to happen. n8n keeps the credentials, API calls, webhooks, and boring execution steps. So when OpenClaw wants to do something sensitive, it does not get direct access to the API key. It calls an n8n webhook. Then the workflow stops. n8n sends me a form: Approve this API call? Here is what the agent wants to do. Here is the payload. Here is the endpoint. If I approve, the Wait node continues and n8n makes the request. If I ignore it, nothing happens. That one step changes a lot. The agent can still move fast. The credentials stay in n8n. The risky action is visible before it runs. And the approval step is just a normal workflow node I can reuse anywhere. Refunds. Billing changes. Posting content. User exports. Database writes. OpenClaw handles the reasoning. n8n handles the execution. The Wait node is where I stay in control. If you are giving agents access to real business systems, this is the kind of boring guardrail you want around them.

@dalethomas50 @MichaelHyatt @Paperclip_AI @NetworkChuck @AlexFinn @gregisenberg @dotta Try it with OpenRouter. Priced at $0,435 /M input tokens and $0,87 /M output tokens => For a model in this category, it's hard to find anything that comes close in the capability/cost ratio.

@Slazerslaze @MichaelHyatt @Paperclip_AI @NetworkChuck @AlexFinn @gregisenberg @dotta I need to try DeepSeek V4 Pro, hearing great things.

Now I’m thinking about jumping over Hermes and installing @Paperclip_AI. I can always use Hermes as an agent. I love the vision of AI orchestration at this level. cc: @NetworkChuck, @AlexFinn, @gregisenberg, @dotta

@MichaelHyatt @dalethomas50 @Paperclip_AI @NetworkChuck @AlexFinn @gregisenberg @dotta With Hermes Agent, I'm pretty happy with DeepSeek V4 Pro via OpenRouter at the moment. Also used Kimi K 2.6, which was also solid.

@dalethomas50 @Slazerslaze @Paperclip_AI @NetworkChuck @AlexFinn @gregisenberg @dotta I’m hoping I can control that in the model selection. If there’s one thing I learned from my OpenClaw experiment, don’t use the Claude 4.7 API for everything!

@MichaelHyatt @Paperclip_AI @NetworkChuck @AlexFinn @gregisenberg @dotta Yes, absolutely. I tried it with my ChatGPT Plus subscription back then. Usage was sucked empty in one day 😜. But that was pre GPT 5.5. So it might have been a model and skill issue on my side.

@Slazerslaze @Paperclip_AI @NetworkChuck @AlexFinn @gregisenberg @dotta It likely depends on which models you use. From my OpenClaw experiment, I learned to use ChatGPT via OAuth rather than Claude via API. The difference in cost was night and day.

@benungeskriptet @EskenSaskia Wenn @EskenSaskia zusagen würde, wäre das mal die Art von Diskurs, den unsere Demokratie und Debattenkultur bitter nötig hätte. Meinungsvielfalt und Meinungsaustausch ist gefühlt erst wieder durch @benungeskriptet wieder in die richtige Richtung unterwegs.

Freundliche Einladung an @EskenSaskia 😊 Ich denke, das wäre eine spannende Diskussion. Besser als in den Kommentaren auf Social Media.

@david_attisaas I just wish the charma feedback loop would be much quicker…

@Slazerslaze nicely said

another reminder the market is dumber than you think it is

@JacobKGilbert @NetworkChuck @MichaelHyatt @Hermes_agentAI @gregisenberg @AlexFinn 😂 yes, it's a hard sell to convince the wife that this is absolutely necessary => but you could use your Hermes to give you some convincing arguments why this benefits (you and all) mankind ;-).

@Slazerslaze @NetworkChuck @MichaelHyatt @Hermes_agentAI @gregisenberg @AlexFinn Sweet! Now to convince my wife that I need a SuperGrok subscription to connect to it. 😅

I’m about to install @Hermes_agentAI. Should I install on local, dedicated machine or VPS? Why? cc: @gregisenberg, @AlexFinn, @NetworkChuck

@rileybrown Why not a dockerized Hermes implementation with access to CC and Codex?

Claude Code & Codex in a sandbox > Hermes

@SimonHoiberg Same here. It is shifting into more high level thinking => which is a good thing. I remember being completely empty during Bug-Fixing sessions, which completely destroyed creativity for that day. As an autodidactic programmer, AI really closes the gap to those 10000 hours guys.

When I built the MVP for Aidbase in beginning of 2024, I coded for 2+ months straight until my arms were falling off. I had done these coding marathons many times before, but this was the last time I did it. Since then, it quickly shifted: - +75% code auto-complete - Cline/Copilot coding by itself - Copilot/Cursor coding in the cloud - Fully delegated feature development I actually don't remember when I last wrote a line of code. Software engineering isn't dying. But coding surely did. x.com/levelsio/statu…

@JacobKGilbert @NetworkChuck @MichaelHyatt @Hermes_agentAI @gregisenberg @AlexFinn With at least 4gb Ram => absolutely

@NetworkChuck @MichaelHyatt @Hermes_agentAI @gregisenberg @AlexFinn Could a Raspberry Pi 5 run it well?

@marcfriedrich Mein Job auf jeden Fall => und deshalb beschäftige ich mich massive mit Agentic Coding um meine eigenen Fähigkeiten mit KI zu erweitern.

AI will change everything You think your job is safe?

@mark_k @giordanorandone Effective Altruism always rings the Sam Bankman-Fried alert in my head.

@giordanorandone I very much disagree with their Effective Altruism roots and I think their constant spreading of doomer messages is hurting the entire AI industry.

I'm getting good interactions with the teams at xAI, Cursor, OpenAI, and Google. They are all very receptive to feedback, which is genuinely great to see. The only AI lab ignoring me is Anthropic, and that's fine because I don't use Anthropic products, and don't plan to. 😏

@LinusEkenstam 😂 Waymoses made my day.

Waymoses.