Michael Versluis🪄

@PhantomStays Haha dit bericht las ik een aantal weken terug ook

Just got off a call with a Dutch 🇳🇱 ecom founder duo running a Shopify store at $12M/year, scaling hard toward $20M+, spending $15k a day on Meta ads, and about to go full digital nomad across Spain, Italy, Bali, and the US. Here's the wild part: every single dollar of that ad spend is earning absolutely nothing. His Meta account bills out of Dublin, so his US Revolut and Slash setup generates zero cash back, zero points. $15k a day. Completely dead. He'd also canceled his Dutch Amex over high fees, so there's no rewards stack anywhere in sight. And on top of that, he's still a Dutch tax resident running a US LLC, which means his profits are sitting exposed to serious Dutch tax liability with year-end approaching fast. We mapped out the full fix today. The immediate priority is stopping the bleed. By shifting the European ad account to bill through a US entity, he can route that $15k daily spend onto a Chase Ink card and start capturing 3x points or 3% cash back on global ad spend, uncapped, regardless of currency. At that volume that's millions of points per year, enough to fund his entire nomad lifestyle across Europe and Bali without spending a dollar on flights or hotels. To build the Chase relationship, he can either fly into New York for a bank run and sit down with Chase, Bank of America, and US Bank in person, or we handle it entirely remotely at Phantom. Either way, this gets moving fast. For cash flow at scale, we're bringing in fintech credit cards the moment his ITIN lands in two weeks. Because their books are pristine, they underwrite purely on biz performance, no personal credit check, no personal guarantee. That unlocks a $300k to $400k monthly credit line almost immediately, giving him the runway to scale comfortably past $20M without sweating payment cycles. Then the structure: 🔐 US LLC continues handling all business cash flow and keeps the banking and credit stack intact 🔐 Panama personal tax residency established for 0% income tax and full KYC bankability globally. Panama only requires roughly two days of physical presence per year, which fits perfectly with his 9-month travel schedule 🔐 Panama Foundation layered on top as the wealth and estate planning vehicle. Total anonymity, ironclad asset protection, and his profits fully shielded from Dutch tax authorities before the calendar flips 🔥 Outcome: a $300k+ monthly credit stack that scales with the business, millions of points banking every year to cover flights and hotels across his new global life entirely for free, and a legal structure that means the Dutch government never sees his wealth again Running $15k a day in ad spend and capturing nothing from it is genuinely painful to look at. The points alone could fund years of first-class travel. That changes now.

@gen_z_mind @MatthewBerman @trq212 This is the thing, the token burning on visuals by Claude is massive

I’m not fully convinced that replacing generated Markdown files with HTML files is the right default when working with AI. HTML can definitely look nicer and provide richer formatting, but it may also increase token usage and complexity without always adding proportional value. In many cases, Markdown is simpler, easier to review, easier to diff, and more efficient for both humans and AI tools. To me, this feels similar to telling console-focused developers or system administrators who are comfortable with vim or nano that they should move to a full IDE just because it has AI capabilities. The IDE may be powerful, but it is not automatically the better tool for every workflow. A better approach might be to keep Markdown as the default output format and generate HTML only when there is a clear need for presentation, styling, interactivity, or final publishing.

Imagine writing an article on X and changing an entire industry instantly. @trq212 has powerful ideas.

@NousResearch @ErikVoorhees When I was reading the lines I already knew it was Hermes, happens sometimes

@ErikVoorhees Hermes Agent

How do I solve this bullshit

@outsource_ Watch out @outsource_ , crypto can make you but also break you. Keep focus on building 🚀🚀

Guys very bullish on $hermesworld > 200+ signups > 20k page visits >v1 in development for release >v0 is currently out for testing Will be doing a few things today to support the token.

i'm using hermes with my employee's but how do i know what they are doing with it? @Teknium

@icanvardar Dickpic?

prove me you're not an ai

@outsource_ Nice and each agent has a own memory with 1 orchestrator? I install this 2 days ago but was searching the kanban that was on the main desktop app.

@MichaelVersluis Use Hermes-workspace.com swarm

Forget about agent orchestration, get your agent to orchestrate computers. My Macbook pro 32GB M1 Pro is where Aurora lives PC1 = 4090 Pair the two together and we have a working formula to build Hermes-World.ai completely with AI. Documenting the journey for you guys 1 step at a time!!

@outsource_ Im slowly getting to that set up, I run multiple companies but in the mean time trying to automate everything, hard to do togheter but my Hermes is going the good way

@MichaelVersluis Yes 1 main orchestrator 10b + tokens

@outsource_ And do you have 1 orchestrator ?

@MichaelVersluis I have multiple agents on all devices, I acually have access to 4 PCS 4090/ 4070 / 3070 / 2070 all on tailscale ssh access

@JakeKAllDay @MatthewBerman No @JakeKAllDay but households need to adapt because their government says.

@MatthewBerman I think you massively overestimate the interest in high tech for 90%+ of households lol

homes will come with built-in compute like central air, heating, solar, etc.

@icanvardar Bro there are no secrets anymore

DO NOT SHARE YOUR SECRETS WITH AI!

@RXed_EU @vmiss33 @Tailscale @outsource_ Hé, the idea to use cowork is good! Tnx!

@vmiss33 @Tailscale @outsource_ Not on Hermes , but with Co-work works perfectly. So will work perfectly with Hermes… You don’t need Mac Book Pro M5 128Gb. You just need Mac Studio M3 Ultra 512Gb and an iPad with @Tailscale , done

OK who's added @Tailscale to use Hermes Workspace on their phone? @outsource_

@outsource_ @vmiss33 @Tailscale Hahaha I’m serious in this moment know

@vmiss33 @Tailscale Meee ✅ gotta fix the dashboard mobile real quick for settings wheel

@dontlockmyacc0 @durov @gentlemanonton Bro..

If you want $TON to be different than the other chain, you as a founder needs to take care of the meme projects seriously because that is the main entrance to bring liquidity. I would say start with the mascot $MAN @gentlemanonton I am thinking how we can leverage AI agent for this mascot? Embed a certain personality and make it alive? dexscreener.com/ton/EQB91RhU3K…

Fees in TON have dropped 6× — to nearly zero. Next step — Telegram replaces the TON Foundation as the driving force behind TON and becomes its largest validator. The focus shifts to tech superiority. New ton.org, new dev tools, new performance upgrades. Timeline: 2-3 weeks.

@kimmonismus The problem is not the speed, the problem is the extending grid can’t take all the incoming energy because it’s not made voor such amounts of energy traffic trough it

80 panels per hour or one every 45 seconds. 1,920 per day. I love it. This is something we should scale up massively: affordable energy, which is necessary for everyone.

@CharlesMullins2 There will never be free energy for the normal people, there is simply to much money in the game for that

🚨 BREAKING: We’re now filming fusion plasma at 100 million °C in real time. This isn’t CGI. This is inside the ST40 fusion reactor. At temperatures hotter than the core of the Sun, matter becomes plasma a state where atoms are ripped apart. And for the first time… We can actually see it evolve. This is the same process that powers stars. If we can control it: Unlimited clean energy No carbon emissions Virtually endless fuel The future of energy isn’t theoretical anymore. It’s glowing… right in front of us. What do you think Will fusion solve energy in our lifetime? Follow me I break down the physics behind the biggest breakthroughs.

@Teknium Even on Sundays the team is shipping, 🚀🚀🚀

Now supported in Hermes Agent! See the PR for info: github.com/NousResearch/h…

@keizervan020 @patricksavalle Lokaal is prima maar bij verre niet in de buurt van een Claude. Claude max abbo kan je al gigantisch veel mee en api kosten zijn ook weer afhankelijk van hoe groot de context is die je gebruikt, hoe groter hoe duurder.

@patricksavalle Hoi Patrick, mag ik vragen hoeveel je ongeveer per maand kwijt bent aan Claude? Ik draai zelf een lokaal model en loop nu wel tegen de context length aan merk ik, op zich met een goed systeem 32GB RTX 5090 en 64GB geheugen.

Ik heb wat hele bizarre AI software design SKILL's die ik nog nergens zag toegevoegd aan de set. Zoals: defect-shift-left github.com/l-gevity/l-gev… 'Places every error detection at the earliest stage of the SWD pipeline that is technically capable of catching it.' Zorgt dat fouten zo vroeg mogelijk in het hele traject worden gevonden. Deze gaat je HEEL veel geld/tijd besparen. Vooral op grotere projecten. 'Analyze my project using the defect-shift-left skill' is de makkelijkste manier om te activeren. Hier is de hele set: github.com/l-gevity/l-gev…

@levelsio @Dyson

I still haven't solved the CO2 bedroom challenge You open the window and you wake up from a 6am garbage truck or barking dogs and sunlight You close it, you suffocate in 1200 ppl at 5am I guess you really need some mini tube in your wall with a vent that opens and closed based on internal CO2 but how do I build that?