
Trying to be an AI Deployed Entrepreneur, whatever that means. chief execution guy - @gethalfbaked
Join 175k others 👉🏻
Joined November 2024
- Tweets546
- Following346
- Followers511
- Likes356
EXTREMELY impressed by this.
An IG influencer bro said Claude Code + Lemo Opuscar made these insane animations in 10 minutes.
I was skeptical so I tested it this morning.
I gave it a script and our brand identity doc.
It took 56 minutes for me to generate 3 edits. The first attached below and the other 2 in the first comment.
About 9.5M tokens on Opus 5.5, so cost-wise not too bad.
100% trying this going forward for Half Baked.
If you want me to send you the link to try it yourself comment “video” below
1
1
93
You can now clone any app with Claude Code and a viral REA GitHub repo, so you never pay for a subscription again.
Apparently.
We spend $50K+ on subscriptions at Half Baked, so I tested what I could actually cancel.
First up: @meetgranola
Why Granola? Seems like the most straightforward to replicate.
Here's the stack I used:
> Claude Code (the build)
> REA github repo (worked out how Granola works)
Building experience
> 1 hour 1 minute of back and forth
> $22.30 at API rates
> Got me 80% of the way there
> No account. No subscription.
> Added a shortcut that starts a meeting in one press [CTRL+OPT+M]
Results
> Used it for a day
> Continued to break
> Transcription kept messing up
> Maybe it's my pea brain but the optimisations I needed to make required me to sign up for tools like Assembly AI which costs $$$
The time I spent fixing it likely cost me more in time than the monthly Granola subscription...
Building it was the cheap part, but keeping it working is what I'm happy to pay for.
Cancelled: 0. Kept: 1 (🫶🏼 @meetgranola)
Would recommend trying out the REA repo, it's very cool. Pics + links to everything I used are below.
10
7
399
Everyone is editing their videos with Claude now.
Apparently.
I tested whether the hype translates into something we’d actually use in our business.
For context, we spend $50K-$100K a month on paid ads which includes a lot of video.
Here’s the stack I used:
> Claude Code (Opus 5.5 high)
> Whisper (transcript)
> ffmpeg (cuts and audio)
> Remotion (the edit)
> HyperFrames (graphics and music).
I also gave it some video inspo links + our brand guideline doc.
Summary
> Original raw footage was 1:23 with a lot of dead space and redos
> It built me a review page showing every cut it made which was cool
> Took ~20 minutes to generate the first cut
> By the 4th version it was on par with our editor’s cut.
> Entire process took 40 minutes from first prompt to final export
> 11% of my 5-hour usage allowance (~$17.29 at API rates)
Would I use it for high-volume production yet? I’d want to test a batch first but I think it’d hit limits quickly.
Worth the hype? 🟢 🟢 🟢 🔴 🔴
Will I be doing this again? Abso-bloody-lutely.
Here’s what I’ll do next time to make it faster and cheaper
> style first
> create a better one shot prompt
> give batched feedback
> save recurring fixes to project instructions
> give it one clear reference for pacing and style
film for the edit
The before/after video + a one-page breakdown of how I did it are below.
12
4
288
I run a 7-figure newsletter business, so I've got a few thoughts on this.
Saying "newsletters are dead" is a bit of a stretch.
AI is going to make content quality an even bigger differentiator. Content that's easy to copy will become less valuable. But great curation and creation will still have a place.
Just because AI starts filtering people's inboxes doesn't mean they'll stop reading things they actually enjoy. People don't always want the quickest way to get information. They subscribe to people and brands and an AI summary isn't going to replace that.
Don't conflate (1) email as a way to reach people (2) newsletters as a type of content and (3) newsletters as a business. They are not the same and AI might disrupt one without killing the others.
It's fair to say that having someone's email address could become less valuable over time. But there's a big difference between people using AI and people letting AI decide what they read. I think that shift will take longer than many in tech expect.
Life is going to be much harder for mediocre newsletters, but the ones producing something genuinely worth reading will survive.
a big take I've been noodling on and I’m ready to call:
the newsletter is dead
@every's core business is a newsletter with 120k+ subscribers so I don't say this lightly.
persistent agents mean i dont look at my inbox
with dots (GrokBots, Muse, etc) I almost never look at my inbox. if I do it's for a specific thing I need to reply to.
i think in a year I'll be doing all of my email this way - fully filtered by my dot / something like it.
this is not new: it's a continuation of the trend of having a “newsletter” folder in Gmail, but 10x more filtered.
what it means
It doesn’t mean people won't read newsletters, but it does mean that getting a person’s email is no longer the guarantee that you have a relationship with them that it used to.
luckily, with persistent agents (and AI coding more generally) there is an entirely new playing field to build relationships with readers. instead of fighting to get an email address, publishers and creators will need to make sure we're installed in our reader's agent harness of choice, so that they get directly notified when we do something new.
we're already starting to experiment with this @every with our MCP and the frontier alerts we send users of our slack agent (every.to/agent)
it's also more affordable than ever for publishers and creators to build their own platform ecosystem: you can have your own website, app etc. and you can use AI to publish 100x more high-quality stuff to your community—so we may see a world where many creators themselves are running their own little empires that have their own apps, feeds etc.
to be clear: It'll take a while before this actually happens, but the idea of getting someone's email creating a direct relationship with them is over.
here's a set of predictions I made 3 years ago about AI and media that panned out pretty well: every.to/chain-of-thought/we…
2
96
The barrier to entry is now the barrier between your ears.
Most businesses don’t need just any software. They need a way to turn how they actually work into software
That’s what @emergentlabs is becoming: the operating system for businesses
Today we’re announcing our $130M Series C at a $1.5B valuation to put that future in their hands
1
3
539
Golden opportunity
Startup idea: Timeleft for Running
Timeleft went $0 to $10M ARR getting strangers to meet for dinner. Same mechanic, but burning calories instead of consuming them.
How it works:
• Matches groups of 4-5 runners by pace, location and availability
• You get matched, you meet, you run
• New runners every week, keep the friendships you make
• Training plans and race goals baked in for retention
Run clubs already did your user acquisition. You're just solving their overflow problem.
Steal it.
279
If you’re looking for something to start, we have every idea we’ve ever published (750+) in one searchable Google Sheet.
Comment Half Baked and I’ll dm you the link
Spread the love
🙌 Congrats to the whole team on getting this out the door @NidheeshKavalan.
Excited to see where it goes next! 🔥
178
















