Grok Bot 1.0 is out 🤖
My bot team already hands work to each other: SEO bot writes, compliance bot checks, I approve, it ships ✅
Introducing Grok Bot 1.0. AI teammates that work alongside your whole team.
Share Bots, collaborate with them in Slack and Teams, invite them to meetings, and watch them learn and become more helpful.
6
ChatGPT is turning into a storefront 🛒
If people ask AI about your business before they visit your site, your product pages and FAQs are what it answers from.
Now in beta: Businesses in ChatGPT.
Claim your business, manage your information, and get insights on how people engage with you.
Eligible businesses can also build a Business Agent to answer customers’ questions and help them explore what you offer.
openai.com/index/introducing…
6
The part of Grok Bot that surprised me: the bots talk to each other.
One writes SEO content for my SaaS. A compliance bot checks it — and my e-com shop, and the ads. Then the update goes out. I'm not copying text between chats anymore.
They've also been making images, shipping page updates.
I still approve what goes live. That's the bit I don't want to give up.
1
11
Whiteboards, sheets and data stories right in the chat 👀
AI is moving from "answer my question" to "build me the working thing".
My Grok Bot team already runs this way: plans and trackers in one Notion hub that every bot reads and updates. I just approve 🔁
Starting to roll out today: Claude Whiteboards, Claude Sheets and Claude Data.
Think a plan through on a board, build a tracker in a sheet, and tell a story with your data.
53
Now the @bot just talk to each other.
✍️ One writes SEO content for my SaaS
🔍 Another checks it before anything goes live — same one watches my e-com shop and ads
✅ When it's clear, the update gets shipped
They've been writing pages, making images, checking the shop, helping with my CV and job search.
I still say yes before anything is public.
I can just spam ideas to CEO Bot who documents all, and does the work..
88
🚨 Anthropic just dropped Claude Fable 5.2 and Mythos 5.2, built for the most ambitious technical work.
But look at their own advice 👇
"Start with Opus 5.5, our strongest daily driver, and switch only as required."
💡 Same lesson I learned with my Grok Bot team: the biggest model doesn't fix vague asks.
✅ Small, clear tasks
✅ One job per bot
✅ Go bigger only when the task needs it
Which one are you trying first? 🤔
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.
They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
31
🚀 My Grok Bot team shipped 4 updates to my SaaS today.
💸 Expensive lesson: "fix the whole site" burns tokens like crazy.
✅ Now it's one fix per PR, and nothing goes live until I approve it.
Small asks ship faster.
3
3
84
🗂️ My setup for running several projects with Grok Bot: one Notion hub with
📌 Projects (status + next action)
💡 Ideas (captured, not started)
🔑 Accounts (where things live)
All my bots read the same source, so I stop re-explaining context in every chat.
1
2
26
My Notion setup separates projects, ideas and accounts. It gives my bot team a place to find context.
The limitation: a written plan isn't completed work. I need to keep ideas, next actions and actual results separate.
1
11
My current setup: 1 primary CEO AI bot + 9 project bots.
→ X news account
→ Shopify store
→ Ads & shop compliance
→ Research
→ Strategy
→ Planning
→ YouTube Shorts testing
→ Job hunting
All documented in Notion, so every bot has context.
Just delegation…. For now
2
31
AI bots help me with the boring stuff now:
research, compliance checks, draft content, status updates.
They don’t help me avoid the hard part — getting traffic and talking to customers.
6
Someone on Reddit made a game where you can ride a motorcycle on the stock chart of any company. xitter.twitterapi.xyz/realCaigu/status/20648…
The duality of tech in 2026: Top: "I used AI to build my own operating system." Bottom: "Please wait 24 hours to generate another div." 💀
Used Claude Design and Opus 4.7 to design a personal dashboard.
With Personal AI becoming more accessible in 2026, I always envisaged that one day we could have our own personal OS as well.
Here's what I came up with!
1
97
The jump in agentic coding and autonomous verification here is staggering.
We are officially in the era where models don't just write boilerplate; they run complex, multi-step workflows with minimal supervision. 🛠️🧠
#AI #SaaS #BuildInPublic #Agents #Claude
Introducing Claude Opus 4.7, our most capable Opus model yet.
It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back.
You can hand off your hardest work with less supervision.
ALT Claude Opus 4.7 Benchmarks
1
1
97
Tired of the “just split evenly” fight at the end of dinner?
Science says: when groups split the bill equally, everyone orders 37% more on average.
It’s called the “Unscrupulous Diner’s Dilemma” — Dave orders the $87 steak because he only pays 1/6th of it. Everyone else over-orders too. Total chaos.
I got so fed up I built whoatewhat.xyz: Snap a photo of the receipt → AI reads every item → you tag who had what → instant fair split. No more arguments.
Who’s been burned by this exact situation? Drop your worst “Dave story” below 👇
whoatewhat.xyz (free)
1
33