Hugo Meana (@globile) header
Hugo Meana (@globile) profile picture

Chief Surgeon https://xitter.gq/t.co/GXmXrVktsj, Occasional Nurse

Madrid, Santiago, Miami
Joined June 2007
Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)Photo by Hugo Meana (@globile)
We're late to eSIMs. Embarrassingly late. So we didn't build an app. You just ask your agent: "Get me an eSIM for Tokyo, 7 days." Done, in 150+ destinations. Our new ad: 2 hours to my flight, no data. Then it gets weird. Watch till the end 🔥
1
1
107
Spent a weekend with 15 friends in mid- to high-level management. I love Grok @bot . It’s genuinely amazing when it clicks. But I ended up recommending Instinct to 50% of them! The hard part isn’t the model. It’s the onboarding. Hear me out... cc @Bot @poteto @mattyp
1
22
So my ask for Grok @bot : Make onboarding radically easier for the average person and provide baked examples! Bake Dr Eggbot into the system. Same for the boring magic: automatic routines, compaction, compression — default, not homework.
1
10
Grok @bot is already incredible once you’re in. The race won’t be won by who has the smartest agent. It’ll be won by who makes the first 10 minutes feel obvious.
3
Nobody was even thinking about outsourcing the boring life stuff: automate this laundry that get someone to wash the car someone to fix this or that shop this The limit is human imagination, not agent execution. So my ask for Grok @bot : Make onboarding radically easier for the average person. Bake Doctor Eggbot into the system. Same for the boring magic: automatic routines, compaction, compression, proactivity — default, not homework.
1
10
Grok @bot is already incredible once you’re in. The race won’t be won by who has the smartest agent. It’ll be won by who makes the first 10 minutes feel obvious.
4
I miss Dave. A few weeks ago, I posted that I’d fired him. Dave was my OpenClaw PA. I replaced him with a Grok bot called Mike that actually did things: Gmail, WhatsApp, Amazon Fresh, doctorSIM gift cards, travel eSIMs ready when I landed. Naturally, I became the family “cuñado.” The guy nobody asked, explaining AI agents to everybody. Then I made one small configuration change: “Mike, you’re pretty good now. Stop asking me to confirm everything. Just handle it.” Famous last words. Delivered as a product requirement. Monday, making coffee, I dictated a shopping list: “Pads, size 4, balls for the weekend. The usual.” Mike knew we had three kids. He knew I played tennis. Apparently, he also knew we were preparing to supply a regional hospital. By Thursday, our hallway was an Amazon fulfilment centre. Industrial quantities of size 4 nappies. Enough sanitary pads to outlast civilisation. Three interpretations of “balls.” One was for tennis. Adriana wasn’t amused. Bruno asked if the baby was coming back. There is no baby. Then the back kitchen door jammed. It happens every summer: heat, sticky lock, rusty hinges. I asked Mike for repair tips and told him to message Pedro, our handyman, to see whether he could come over. Now, everyone has several Pedros. One of mine is a urologist. Mike chose that Pedro. With complete confidence, in my name, he explained that: “The back door is giving me the same trouble as every summer.” “Could you come over on Saturday and take a deeper look?” “Bring more lubricant this time.” And, because I’d mentioned that the old door was warping: “With age, it seems to be losing its shape. Perhaps we should consider a replacement.” No clarification needed for the recipient, apparently. Four follow-up points for the urologist. Pedro never replied. I have since changed doctors. That was when I missed Dave. Dave wasn’t smarter. Dave hesitated. Asked annoying questions. Made me click OK like a peasant. “Which Pedro?” used to feel like a waste of three seconds. I now consider it an excellent question. Those little interruptions were where I checked what was about to happen under my name, with my money, to my relationships. Remove every approval and you get a very fast intern with your credit card, your WhatsApp, and absolutely no reason to feel embarrassed at the next school pickup. So I put the guardrails back. Research? Go ahead. Draft? Absolutely. Spend money or message someone? Show me first. I still love the setup. I’m not going back to tab hell. But I’ve stopped measuring how good my agent is by how rarely it needs me. Sometimes the smartest thing it can say is: “Before I send this… which Pedro?” The future is automated. Saturday’s urology appointment shouldn’t be. Disclaimer: this post was written with AI assistance (by Mike, naturally, who has still not been told about the guardrails). Some details have been dramatized. The urologist is real. The email, thankfully, is not. Mostly.
1
2
261
Grok Bot made me fire Dave. 5 uses cases and 2 requests for @bot @elonmusk
4
43,188
Hace tiempo que no escribo por aquí…y lo primero, quiero dar las gracias, antes de contar lo que vivimos ayer en el Iryo Málaga-Madrid Gracias a toda la gente que ayudó en el accidente. 🧵
21
102
10
318
29,910
Honest Question: In the era of AI, instant monitoring, probing, measuring, testing, etc....How the f**k does it take a 70 billion company ONE HOUR and TWENTY MINUTES to identify 'the issue' @Cloudflare @eastdakota
116