SOME GROK BOT USERS LET THEIR BOT PAGE THEM WHEN SOMETHING IS URGENT. THE EXACT RULE I GAVE MINE SO IT NEVER WAKES ME FOR NOTHING.
grok bot's product lead calls the v1 setup simple. the bot sits on top of slack and email, you tell it your role and what you care about:
"i want you to notify me in these cases. in these cases, you don't need to ping me directly, but you should include this in your daily roundup"
some users went further and let it page them. his warning: "you really want to trust that grokbot does not have false positives"
my rule, 3 levels:
→ page me: a login i didn't make, money leaving my accounts, my bank or my doctor
→ text me: anything that needs an answer today
→ 6pm roundup: everything else
→ never a page for newsletters, sales or "quick question"
the prompt:
"read my email and slack. page me only for a login i didn't make, money leaving my accounts, or my bank or doctor. text me what needs an answer today. everything else goes in a 6pm roundup. if you're not sure, it's not a page."
a login i didn't make is the page i care about most. how grok bot guards my accounts, article below ↓
THE GUY WHO BUILT GROK BOT SAYS HIS FIRST PROMPT WAS ONE SENTENCE. IT FOUND 2 JOBS NO AI COULD DO BEFORE.
his advice for day one: give it "the context it needs to be successful", like onboarding a new hire. slack, email, the records you use every day.
then his first task, no setup screens:
"go through my slack, go through my email and suggest like five things that you can take off of my plate"
it suggested five. two were really useful. he spun off a bot for each on the spot.
"it was not like draft an email, it was like do a chunk of work"
how to copy it:
1. connect email, chat and calendar first
2. send that exact prompt
3. ignore the weak ideas, keep the 2 good ones
4. one bot per job, not one bot for everything
5. power tip from him: make every bot write to one place, a daily digest you read in 2 minutes
if it reads your email, lock the email first. grok bot security setup, article below ↓
ELON MUSK: WHEN ENGINEERS AT ANTHROPIC AND OPENAI SAY THEIR OWN AI IS DANGEROUS, "WE SHOULD BELIEVE THEM."
he was asked what he meant by "Dario is right."
his answer:
"the danger of AI is very significant at this point"
"we have exponentially increasing risk with the AI models"
and the people saying it build the models themselves.
"when a lot of people from Anthropic and from OpenAI are telling you that their models are very dangerous, I think we should believe them."
then he pointed at the irony: a "10% chance of annihilating humanity. but by the way, how much allocation would you like in our IPO?"
"that's some crazy 4D chess."
if the builders are worried, check what your own AI can reach. my grok bot setup is in the article below ↓
ELON MUSK SAYS TO AI, TALKING TO YOU WILL FEEL LIKE TALKING TO A TREE. HIS FIX GOES INTO YOUR HEAD.
his numbers: you put out "maybe 100 bits per second". computers talk at "a trillion bits per second".
"talking to a human will be like talking to a tree."
his fix is neuralink. until then the slow part is your typing, so cut it:
→ talk instead of type. voice is about 3x faster
→ send a screenshot, not a description of the screen
→ drop the whole file in, don't paste pieces
→ keep one context doc about you and your project, attach it every time
→ ask for 3 options and pick one, instead of explaining what you want
less typing from you, more output from it.
i stopped typing passwords too. how grok bot handles my logins, article below ↓
ELON MUSK THINKS AI COULD END US. HE'S BUILDING IT ANYWAY. HIS REASON TAKES 10 SECONDS.
isaacson: sometimes you call it an existential threat, now you go with the flow.
elon:
"some percentage chance of it being an existential threat."
"there's nothing I can do to stop it."
"either be a participant or a spectator."
"being a participant is more interesting than being a spectator."
and the last word of the answer is "maybe."
can't stop it either. so i put it on guard duty for my passwords. setup in the article below ↓
f*ck. elon musk says he's had AI nightmares "many days in a row." then he explained why work will be optional
"I've had AI nightmares many days in a row."
"what am I supposed to do about it?"
then the forecast.
robots and AI do anything we want.
"work will be optional."
that's not what he wants.
"if I could, I would certainly slow down AI and robotics, but I can't."
nobody can slow it down. so mine guards my accounts. article below ↓