Yanick retweeted
Are people forgetting this was already possible and much better?
This is just Google making it easy for consumers to use with a sleek interface.
-
Today, we’re introducing Pomelli’s latest feature update, ‘Photoshoot’
With Photoshoot, you can start from a single image of your product and easily create high quality, customized product shots to elevate your marketing.
Available free of charge in the US, Canada, Australia & New Zealand! Get started with Pomelli today at labs.google/pomelli
Yanick retweeted
How AG1 replaced a $221K/year UGC team with this Nano Banana Ad factory.
Most brands are stuck in the UGC nightmare:
Paying $15K+ monthly for inconsistent
Creators
Waiting 2-3 weeks for basic testimonial videos
Getting 5-8 generic videos max per campaign
Creators ghosting mid-production
This system obliterates that completely.
Upload your product → AI generates 200+ UGC variations → ready-to-launch video ads.
Not mockups. Real scroll-stopping content.
Built after watching brands burn $200K+ on creator networks that deliver recycled testimonials.
The efficiency gains? Ridiculous:
→ 20x UGC production velocity
→ 97% cost reduction vs agencies
→ Unlimited video variations
→ Zero creator dependencies ever again
How it works:
AI Avatar Generation Engine, Creates 15+ authentic spokespeople in seconds
Script Intelligence System, Maps product benefits to testimonial angles automatically
UGC Production Stack, Generates lifestyle demos, unboxings, testimonials instantly
Platform Optimization, Auto-formats for TikTok/FB/Instagram algorithms
47-second production time per video.
Complete UGC independence.
Yanick retweeted
I find every ai saas landing page looking exactly the same.
- Simple header
- the circle profile picture testimonials
- The money saved / time saved
- The boring looking buttons
- Simple footer
When are you guys coming up with something new….
Yanick retweeted
What if I told you that this is possible for every small account?
Like and retweet this post and comment “Small”
I will dm you how to target premium users only and that’s how you get paid more with less impressions.
Yanick retweeted
I sold the exact same AI service at 5 different prices.
$97 → nobody wanted it
$4,997 → everyone bought it
The shocking reason why 👇
medium.com/@haertdijkai/the-…
Yanick retweeted
Why Claude Code Exists And Why It Destroys CodeX, Cursor, Replit, etc.
Most think “better AI model = better coding tool.”
Wrong. The system around the model matters more than raw IQ.
Claude Code’s Origin Story:
Anthropic didn’t build it for consumers.
They built it to supercharge their own engineers
shipping features, fixing bugs, and running multi-agent workflows faster than human-only teams could.
That meant:
Every prompt & workflow optimized for Claude’s quirks
Project-completion checks baked into the core
No “MVP” only the features Anthropic devs actually used daily
It wasn’t supposed to be released.
Internal debates happened. Boris (the creator) even said they weren’t sure it should go public because it was that good.
So why release it?
Talent magnet: Show the dev world Anthropic is the place to be
Ecosystem play: Get more engineers building in Claude’s style
Competitive pressure: Cursor, CodeX, Gemini were creeping into the space
Here’s the key difference:
OpenAI’s CodeX = “Rocket engine on a bicycle” huge power, but the frame can’t handle it.
Claude Code = Formula 1 car designed around the engine from day one.
Even GPT-5 with raw intelligence can’t beat a tight, integrated system that was battle-tested inside a billion-dollar company before release.
If you’ve tried Cursor or CodeX and then Claude Code…
You already know.
Yanick retweeted
While Reddit argues GPT-5 is a downgrade, I made $127K in 7 days with it.
No hacks. No team. Just Agent Mode + the right prompts.
Here’s exactly how:
medium.com/@haertdijkai/howc…
Yanick retweeted
GPT-5 is Here… and It’s Underwhelming But These 6 Tweaks Unlock Its Real Power:
I’ve been testing GPT-5 for 4 hours straight.
And honestly? It’s not the leap we all hoped for.
But here’s the twist with a few key tweaks, you can make GPT-5 perform like the upgrade we were all expecting.
Here are 6 practical changes you can apply right now.
1. Unlock “Expert Mode” With Self-Critique Loops
GPT-5’s first answer often isn’t its best.
After it replies, immediately say:
“Critique your answer, find weaknesses, and improve it.”
This triggers a refinement loop that feels like human expert review.
2. Split Roles Into Sub-Agents
Don’t ask GPT-5 to do everything at once.
Break big tasks into
Researcher → Strategist → Writer → Editor.
Removes the “jack of all trades, master of none” effect.
3. Use Output Formats That Force Precision
Tell it to respond in JSON, tables, or step-by-step lists.
Forces GPT-5 to commit to details instead of vague answers.
4. Leverage External Memory
GPT-5 forgets context over long sessions.
Save outputs, feed them back in as a “project log,” and watch consistency skyrocket.
5. Chain-of-Persona Reasoning (New in GPT-5)
Just found this:
GPT-5 can hold multiple personalities in one answer without blending them.
Prompt:
“Answer this as 3 different experts a risk analyst, a visionary founder, and a skeptical investor then merge their views into one final decision.”
GPT-4 always bled the voices together. GPT-5 keeps them razor-sharp and then integrates them like a debate.
6. Real-Time Hypothesis Testing
GPT-5 can now simulate scenarios step-by-step and change its approach midway if evidence shifts.
Prompt:
“Plan this marketing launch, but at each step, pause and ask: ‘If this assumption was wrong, how would I adjust?’”
GPT-4 stayed locked into its first plan. GPT-5 actively course-corrects like a strategist in a live meeting.
Final Take
GPT-5 isn’t broken it’s just under-optimized out of the box.
Apply these 6 tweaks, and you’ll get the leap forward you were expecting.
Have you found a tip that boosts GPT-5’s output? Drop it I’m testing them all.
Yanick retweeted
If you are a content creator and work with many different people, @NotionHQ is a great way to work on a content planner together.
And if you are ready you can even publish it to show to clients.
More people using notion for this?👇🏾
Yanick retweeted
AI SEO is killing traditional SEO.
I made a complete breakdown of how it works
and how to rank higher yourself!
medium.com/@haertdijkai/ai-s…
Yanick retweeted
Why was @Gab__AI banned from X?
Because it wasn’t just another chatbot it was a weaponized AI trained to spread hate, conspiracy theories, and antisemitic rhetoric. Here’s the full story. 🧵
Yanick retweeted
🚨 The $1.8 Trillion AI Secret 🚨
Norway’s wealth fund accidentally uncovered a strategy so powerful it’s worth 213,000 hours of AI compute.
And almost nobody noticed.
👇 Here’s what they stumbled into (and how it could change everything):
medium.com/@haertdijkai/the-…
Yanick retweeted
THROW YOUR XML PROMPTS IN THE TRASH!
We are now experimenting with prompting trough @neuralink
Just think about a prompt and ai will read your mind and give the perfect outcome!
Yanick retweeted
How I got chatGPT to actually sound like me when writing stuff.
Just wanted to share a quick tip that helped me get way better results when using ChatGPT to write stuff in my own voice especially for emails and content that shouldn't sound like a robot wrote it.
I kept telling it “write this in my style” and getting generic, corporate-sounding junk back. Super annoying. Turns out, just saying “my style” isn’t enough ChatGPT doesn’t magically know how you write unless you show it.
Here’s what worked way better:
1. Give it real samples.
I pasted 2–3 emails I actually wrote and said something like:
“Here’s a few examples of how I write. Please analyze the tone, sentence structure, and personality in these. Then, use that exact style to write [whatever thing you need].”
2. Be specific about what makes your style your style.
Do you write short punchy sentences? Use sarcasm? Add little asides in parentheses? Say that. The more you spell it out, the better it gets.
3. If you're using ChatGPT with memory on, even better.
Ask it to remember your style moving forward. You can say:
“This is how I want you to write emails from now on. Keep this as my default writing tone unless I say otherwise.”
Bonus tip:
If you’re into prompts, try something like:
“Act as if you're me. You’ve read my past emails and know my voice. Based on that, write an email to [whoever] about [topic]. Keep it casual/professional/funny/etc., just like I would.”
Anyway, hope this helps someone. Once I started feeding it my own writing and being more clear with instructions, it got way better at sounding like me.
Yanick retweeted
“AI is coming for our jobs”
I would love to see an AI robot sell adderall to its nephew’s friends
Yanick retweeted
Everyone can build.
But just a few understand the way of marketing to a targeted audience.
If you are one of them, it doesn’t matter what niche you build in you will always outplay the majority.













