brian
1.6K posts

brian
@doranalytics
creators are the new engineers
Venice, CA Katılım Ocak 2021
415 Takip Edilen585 Takipçiler

.@mcuban says humanoid robots won't last more than 5-10 years.
Instead, we'll "design the house to fit the robot, and design the robot to fit the house."
"You could create a house where the pantry, the refrigerator, and the washing machines were hidden behind the garage, if you even have a garage. That way you could redesign the house so that all the living space was for people."
From his appearance on the show in March.
English

lots of people building agents from the "workflow" side and not the "personality" side
i pay $10/mo for Tomo, initially just as a calorie tracker to check in with me daily
now i use it as a scheduling assistant for time management, i track my entire day in 30 min increments for accountability (it pings me and i tell it what i did and it backlogs it on calendar) - very helpful to audit how long you spent on stuff
but it also has a cool personality and i use it for travel recommendations, quick web research, weather, marketing ideas, jotting notes, setting reminders
it now allows you to connect your bank, and so it replaced monarch, it can see business cash flow
Poke is another example of this "personality as a moat" AI which i think is greatly under explored
sure claude can do this, but i operate my life out of iMessage, opening an app is slow and clunky
English

after chat, pplxty, and claude
people forget that no AI tool really slapped out of the box for several years
(maybe midjourney as the exception)
but everything was garbo
jasper, lovable, etc
all trash, until years later they "got good"
the bar has been raised now, things like @higgsfield_ai and @origamichat go pretty hard at launch
maybe this is an obvious observation but nothing cuts thru anymore unless it works
(save for the manufactured launches like that horrid ai cmo thing okanda or whatever)
English

maybe im wrong but feel like modern startup advice is anti-Jobsian
@paulg "make something people want"
and
(unknown) "validate quickly and iterate"
both assume people know what they want, and that users are the best product designers
ofc this is great advice in general, but idk if it leads to net new ideas, which is what is so cool ab startups
i want to build something no one thought they needed until they found out it existed, and design it in such a way that there is no feedback bc it just works
English

as long as most tokens are subsidized the reward function will be to limit energy usage, leading to suboptimal responses, such as condensing detailed spec files and the AI playing "telephone" with your instructions leading to downstream slop.
obvi i am super bullish on everything but dont see an end to this in the next decade.
just means more value continues to accrue to the people obsessing over the details, same as it always did
English

who is the @higgsfield_ai head of product marketing?
yall are cracked out of your gourd and deserve some flowers




English

“Distribution is the moat” is kind of a trope now
But I think the bifurcation is more clear
You are either a tremendous engineer turning energy into intelligence
Or you are a media company
There is no in between
It’s Pareto’s principle and you already know where you fall
Really helps frame all those UI, UX, TTV “product” decisions
English

the reason why corporate MBA types are not good at product and marketing (and i am including my former self in this analysis)
is because the incentives are to make things ultra-complicated and nebulous, so as to conjure an aura of authority
the MBA's only inherent value is that it enables the user to communicate with other MBA types in the land of corporate jargon
"never trust a person who refers to Income as EBITDA" says W. Buffet
i think my MBA is a hinderance to being a founder and finding fast PMF
"you have enough ideas to kill Amazon" as one executive told Bezos
MBAs live in the land of the theoretical, founders must be hard-wired to ground truth
English

excited to see more brands/apps emerge and learn that content itself is a marketplace, and AI avatars have no incentive to make better content for you
(save me the RLHF / reward function narrative - it leads to more work for the brand)
rev share / equity / bonuses drive real performance and creates real network effects
(it helps when the person selling your product is capable of using it)
the onus will be back on the brand to "make better creative" realizing the time-for-money arb of AI avatars was a costly detour and they could've just paid human creators more money (and now they cost more)
English

Excited for game two on Wednesday at MSG!
… need to lock down a first three row ticket again — oh the humanity!
#lifeofjcal 😂
English

@volosatovde @tinyaffiliate You should do marketing instead of shipping new features
English

I've shipping an update for @tinyaffiliate
Creating an affiliate program for your SaaS
is now as easy as vibe coding.
English




