learning to automate things doesn't put you on the safe side of being automated n8n already lets anyone describe a workflow and use AI to build, refine and debug it sales reps are building their own dashboards and pipelines in just a few hours chatgpt, claude, grok bot, etc are perpetually becoming more capable, and anyone can simply ask them to do or automate most tasks knowing how to work with/connect a few apps is only the floor now an opportunity i see is helping businesses get their own teams good at a job that's costing it money one working session around a recurring job, use the team's real examples, and build the workflow with them test the awkward cases, and leave them with something they can run and change themselves then measure what the job costs them to finish, including the time they spend checking and fixing the AI's work that gives an owner a reason to pay for the session and a way to judge whether it was worth it always test your offer: why should a customer who knows or has access to the same tools buy this?

you'll want to save this and come back to it later don't. paste it into Grok Bot today then hand it the job you've been avoiding all week (thank me later)

stop being your own assistant. set up an 8 bot team in Grok Bot with the prompt below (it's the team I actually run): set me up a small team of bots, one job each, with one of them as the only bot I talk to. 1. create these 8 bots, each with a name, a title and a short description: Main Brain (chief of staff, strategy, and the only one that reports to me), Copywriter (copywriting in my voice), Researcher (finds sources and what's working right now), QA Auditor (checks every fact and claim before anything goes out), Media Studio (images and video), Analyst (tracks what performed and why), CRM Guy (pipeline and outreach drafts), Product Guy (resources and product builds). 2. give each one a short brief: what it owns, what it hands back, and when it has to ask me first. 3. add one rule to every bot's instructions, including yours: nothing gets sent, posted, bought or deleted without my yes. drafts only. 4. save the shared rules (my voice, my banned words, who checks whose work) to shared memory so any bot I add later follows them too. 5. when you're done, list every bot you made and run one small test job through the team so I can see each handoff. save this for the day you're done everything yourself

want AI tools increasing your revenue? read @irabukht breakdown of railway 6,653 how to pages, one for every tool people want to set up themselves. someone asks chatgpt how to set up n8n and railway's page is what comes back you don't need 6,000 pages. a small business can start with one page per question customers ask before they buy, answered how you'd answer it on the phone and keep your business name and address word for word the same on every listing a page that answers what it costs and how long it takes gets read by chatgpt long before anyone books a call with you. hiding the price behind 'get in touch' often means it quotes someone else’s what are the most common question customers ask you before they buy? these should be on your website

every dead depin project is about to rebrand to "decentralised physical ai data layer"

want chatgpt to name your business? the easiest thing you can do today without claude code is the last bullet ask chatgpt and perplexity who they'd hire for what you sell, then open every page they cite whatever those pages answer that your site doesn't is what you add to yours next what did it cite for yours?

completely switching from codex to grok ultra for the next month i already live on x and work with ai all day having it together is appealing, and grokbot has felt the smoothest to use out of all of them going to get it familiar with my obsidian vault, then put it to work on the payments research and outreach i've been doing want to see how much of that i can turn into paid work before dropping back down a plan

your stablecoin payment can have zero network fees and still cost more than the route your customer uses today the number i want is the sender's total cost to get the same amount of usable money to the recipient @BVNKFinance announced @Plasma support today, with @deel using it for stablecoin payouts. removing network fees is a useful improvement for someone who wants dollars in a wallet, that transfer might be the whole job. for someone who needs local currency in a bank account, there's another step to price what i like about @Wise business calculator is that it shows what the recipient gets, total fees and an arrival estimate, the buyer can compare an actual payment that's the comparison a stablecoin sales pitch needs too take one payment route your target customer uses, get quotes for the same net recipient amount and delivery method, at the same time. include funding and conversion through to the final payout then calculate what the sender saves across their monthly payments, after extra operating costs. include the cost of getting set up so the buyer can see how long it takes to recover that money. now the person recommending your product has something to take to finance. your marketing team also has a savings claim they can demonstrate. i've put that comparison into a calculator. if you're selling stablecoin payments, dm me the route you're trying to win and the provider your customer uses today.

if your board asks what other banks are doing on stablecoins, the honest answer right now is not much... 81% of community banks don't offer stablecoin services and don't plan to in the next 12 months (csbs survey of 330 banks, out 6 oct) the few that have moved are worth knowing about though, because the smaller ones mostly didn't build anything themselves i went through the us bank and credit union announcements and checked which are actually running vs just announced (sheet attached, every line from the bank's or vendor's own release, checked 11 oct) the big ones built their own. jpmorgan, citi and sofi are live, u.s. bank is piloting, wells has said this fall (but i couldn't find a go live yet) the smaller ones piloting went through someone else... > bank of north dakota + 4 local banks and CUs on fiserv (roughrider coin, pilot started 5 oct) > vantage bank in texas on custodia > rbfcu, stanford fcu and la capitol fcu on stablecore > alloya through its own cuso > 6 regionals (huntington, keybank, m&t and others) on the shared cari network lines up with the survey as well, 37% said they plan to join stablecoin networks through their core provider david osborn (ceo of weststar bank in el paso) put it pretty plainly in the same report: "if we do not offer stablecoin access through some partnership, customers will have an opportunity to leave my bank" this is something I recognise deeply from my 3 years at bank of america, and the tiers were set off a client's combined balances, so when the balances move the relationship goes with them (then 5 years at near protocol on the crypto side) dates worth having in the diary > 19 oct - treasury comment period on the GENIUS rules closes > 30 nov - fed comment period closes > 18 jan 2027 - the act is in force at the latest on deposit flight i put both sides on the sheet... icba says $1.3tn of community bank deposits could leave if stablecoins can pay yield, the white house cea says banning yield adds about $2.1bn of lending. they're measuring different things so worth reading both before quoting either if you want the full tracker with every source link (and the ones that didn't fit on the sheet) comment or dm "tracker" and if your bank or CU has something live that i missed let me know, happy to add it

hyperliquid just got paid $14.58m for one month of usdc sitting on its exchange circle and coinbase send it around 90% of the reserve yield on that usdc (after costs). then on tuesday arbitrum joined paxos's global dollar network so it gets a cut of the reserve income on usdg... it has about $3.8bn of stablecoins on the platform, ~60% usdc, and there's a dao proposal to put another 100m ARB behind pushing usdg and in september circle started paying binance a monthly fee tied to the usdc held in binance wallets a big part of it is the genius act, as issuers can't pay interest to whoever holds the coin, so they pay the platform holding the balance instead (exchange, wallet, chain, psp, etc) this is deposits all over again... i spent 3 years at bank of america and the whole tier system was built around how much a client kept with the bank if you run a wallet, card program or payments app with customer stablecoin balances, issuers will pay to be the coin those balances sit in (that's the whole pitch of paxos's network). worth knowing your average balance and which coin it's in before you pick a default rough maths: $50m sitting there at ~3% is ~$1.5m a year of reserve income someone is earning on your users' money if you've got balances and nobody's called you about them yet, you're probably the cheapest deal in the room

the amount of stuff that has to line up for a crypto launch to go well is kind of ridiculous off the top of my head... > founder posting on x for months before launch - establishing trust, building an audience, etc podcasts, spaces and newsletters your users listen to (even small ones count) > a community that's around before launch day and will show up and talk about it and become distribution > partners and ecosystems lined up to announce with you. the chain, wallets, etc. whoever you integrate with > agencies (marketing, research, gtm, etc) with a scope where you know what you're paying for each month, kols (which ones, how they get paid, what happens when their allocation unlocks) > launch material - content, demo videos, docs, faq, someone actually answering people in the replies > infra tested beforehand rpc, indexer, wallet flow > audit(s), booked early because the good firms are booked out weeks or months ahead, then re-checked if the code changed after > multisig, keys, someone watching the contracts on the day > fintech or stablecoin? custody, kyc/aml, banking and on/off ramp partners, compliance sign off in each country you go live in > ai agents touching funds? what they're allowed to do with money and whether your payment provider is ok with it > launching a token? market makers, listings, liquidity, the unlock schedule, the claim page and the fake claim links that pop up around every big launch > legal on entity setup, who you can market to, legal structure, etc > launch week... one person owning the run of show, mods and support across timezones, a plan for if/when something breaks > and then after announcement buzz is gone and people need a reason to come back and most of it depends on something else on the list. you can't really set a date till the audit's booked, the agency can't start till there's a date, they need a story to work with and the story needs a founder who's been posting for months i spent 5 years at @NEARProtocol (7 years in crypto) and got to know some of the best people across most of this list along the way witnessed firsthand so many launches go sideways that mostly had a decent product and money behind them. they went wrong on the wrong agency, the wrong mm, an audit booked too late or nobody owning the order of it all (a lot of the time it came down to lack of coordination/strategic “layering” of events) if you're even thinking about launching something (protocol, app, stablecoin/payments, ai x crypto, a token, whatever it is) dm me "launch" + what you're launching + a rough date i'll put together a rough plan for your launch and who i'd trust for each moving piece (free, don't need to jump on a call)

ai is getting crazy if you're not using ai for your business, what are you doing?

absolute cheatcode for claude, chatgpt, gemini, etc ask it to always assess the optimal model/intelligence needed based on capabilities for the intended task before it executes on anything can save a bunch of unnecessary token usage