22 || https://xitter.cf/t.co/Xj8AiQ9uns, https://xitter.cf/t.co/owdh1F6Cok.
Joined June 2024
- Tweets6.5K
- Following419
- Followers3.4K
- Likes15.3K
you don't need next.js + vercel + supabase anymore
- next.js → tanstack start or vinext
- vercel → workers
- supabase postgres → d1
- supabase auth → better auth on d1
- supabase storage → r2
- realtime → durable objects
- redis → kv
- background jobs → queues + cron triggers
one worker, one bill, ~$5/mo
love nevo but i disagree
the best framework in the ai era is the one with the most built in
not the fastest. not the trendiest
ai writes any language now. pick the one that saves you setup
and spend the time on distribution
Because, because - Laravel.
Yes it’s a good framework very robust, used it for years. but it’s a FRAMEWORK.
I used Laravel, Symphony for years, and also some older frameworks like codeigniter.
None of the people gave a good benefit like: speed, faster deployment, multi threading, lightweight.
The only reason to use it, is if you have an old system.
No reason to use it, especially in the vibe coding era.
Use NestJS.
laravel ships with things the js stack pays 6 services for
- auth → built in
- payments → cashier
- jobs → queues + horizon
- cron → scheduler
- websockets → reverb
- feature flags → pennant
- monitoring → pulse
chatgpt still says next.js + vercel + supabase
that's an AI visibility problem, not a laravel problem
if anyone at laravel wants to fix this, dm me
marc lou alone triggered beehiiv so bad they started shipping a feature a day
#14 a faster dashboard
your dashboard just got faster
the app now ships ~45% less code before your first screen renders, and the analytics cards that used to time out on big publications load much faster
it's the first screen you open every morning. it should feel instant ⚡️
i was considering render after the railway incident
then the ceo showed up to dunk on a competitor's outage
changed my mind
it could happen to anyone. including you @render
this used to happen to me every 2 days
then i got the free oracle vps. most of my stuff runs there now
i use it from my phone, or give it a long task with a bunch of plans and close my laptop for the day
yes, the free one
also chatgpt using 22gb of ram is crazy
because when you install grok bot or muse, connecting stuff takes one click. no hacking it together
most people don't want to build their agent or spend time on it
they just want one that works
it took 16 days of resets to ship one actual improvement
hot take but
if your app is genuinely good you don't need these dark patterns
dangerous combo:
an app people tell their friends about
+ shareable features
+ time
you'll grow slower but you won't need a fake 10 min countdown
Glam AI is no longer available on the US App Store!!
launched in Feb 2021.
est. $550K revenue in Sep 2026.
I went back through its recorded onboarding:
- AI photo and video previews.
- gender and creative-goal questions
- a “Preparing…” checklist
- paywall with a 10-min countdown
I don’t know what happened here, no confirmed reason, but it’s a reminder that years on the store don’t guarantee the next one.
1.2m views -> 23 subscriptions
x is where founders are
linkedin is where marketers are
reddit is where your customers are
don't depend on twitter for marketing unless your product is made for twitter, like superx by @robj3d3
try reddit, linkedin, threads, instagram, tiktok, cold email. anything
I don't think SaaS is dead, but "audience building" is definitely overrated.
This launch post has 1.2M views, and the guy who posted it has 400k followers.
Marc built an alternative to Mailchimp/Beehiiv that costs 83% less. My X timeline is full of opinions about it.
And yet, in 5 days, Marc sold only 23 subscriptions at $19/month.
this is the best example of how twitter is a bubble
most consumers and b2b owners i know still read their emails, ai or not
if you're out of ideas or feel like software is dead, delete twitter for 2 months before the pessimism kills you
The threat to @beehiiv isn't marc lou (guy's a jester).
The actual threat to beehiiv is that humans will stop reading their email.
AI agents don't read newsletters.
Even today, the Apple summary front-runs the need to open for a percentage of readers.
This means the ad model for newsletters is gone in 3 years.
The conversion math also becomes tougher and tougher, even with a huge list.
If everyone is using Claude/Muse/Grok as their LifeHQ (which they will), the only newsletters that get read will be personality/opinion driven (for fun).
The "own your audience" frame also dilutes when the layer between your words and the human buyer is designed to avoid consuming it.
This is not a beehiiv problem, it's an email problem in general.
Beehiiv is diversifying into other products, which is smart.
But one of the core tenets of audience building (just get their email and you're golden) is starting to weaken.
so they drain the usage faster so they can announce a reset every day
genius
There were reports of usage draining faster than anticipated for some users over the weekend. We've identified approximately 40,000 accounts that had an inconsistency on which usage limits where the actual usage that was applied did not correspond to their actual selected plan. We are reaching out individually to all these users and have corrected the issue. If this affected you, we'll add credits to your account. Sorry we got this wrong.
