everyone's asking: hire a fresh grad sdr or buy an ai sdr?
wrong question.
both will message whatever list you hand them.
a bad list worked by a human is slow.
a bad list worked by ai is just faster spam.
decide who you're talking to first.
then pick who presses send. – at Chandigarh, India
a high reply rate with zero interest isn't outbound working.
it's the market politely telling you you picked the wrong segment.
stop celebrating replies.
count meetings with people who can actually buy.
that's the only number we get paid on. – at Chandigarh, India
Kabir Arora retweeted
Introducing Effects in @napkin_ai 🚀
Take any Napkin visual. Transform it into a studio-quality image with different aesthetics. Choose from realistic, editorial, sketch-style, and more – or describe exactly what you want.
The structure stays. The content stays. The feeling is now yours to choose.
Kabir Arora retweeted
Most people believe they are their intentions: what they plan (in the future) to do.
In reality you are what you are currently doing and the future is simply a projection of that.
You are action, not intention. This is sort of a life hack because then you can just spend time on improving action without worrying about intention (that usually solves itself).
for startups in sf -
i cover ai stuff at instagram.com/startkabir
i'll make a free video for your company/product over the next 3 days
send your add in my dm
Kabir Arora retweeted
This morning, I went ahead and put $5,000 directly into the hands of builders.
And honestly, it says a lot about where this ecosystem is heading.
I’ve been deeply involved with @OpenxAINetwork for a long time now, supporting the hackathon program from day one.
Over the course of it, we saw 1,200+ builders and 400+ projects come to life.
But today-during the final round, when prizes were being evaluated and distributed-the quality of projects truly stood out.
Some of these ideas are genuinely competing with multi-million dollar companies like @Lovable , @boltdotnew , and @n8n_io .
What makes it even more impressive is that many of these projects were built by:
Solo developers who are completely cracked devs!
or Sqaud that execute exceptionally well together
That’s when it really hit me the true potential of hackathons.
With the right attention and resources, platforms like these can go on to challenge even the biggest monoliths.
Now what the big companies need to learn..
They’re doing all of this without access to proper GPUs.
Over and over again, I heard the same thing:
“I could’ve chosen a better model if I had access to better GPUs to run it locally.”
And that’s exactly the gap we want to fill.
So we’re announcing $3,500 worth of GPU credits for everyone who made it to the final list.
This is the kind of initiative I’ll always stand behind.
The biggest Pill to win is
Keep building.
Keep shipping.
And keep the pressure on to monoliths whether you’re a solo developer or part of a team.
4/ The worst part? They're creating unrealistic expectations.
"AI will 10x your productivity!"
"Replace your entire team with AI!"
Then people try it, get disappointed, and think AI is useless.
The hype is killing actual adoption.
The AI hype cycle:\n\n1. \"This will change everything!\"\n2. \"This is revolutionary!\"\n3. \"Wait, it doesn't work as advertised\"\n4. \"Actually, it's pretty useful for specific things\"\n\nWe're currently between steps 3 and 4 for most AI tools."
AI safety researchers: \"We need to prevent AI from destroying humanity\"\n\nMeanwhile AI: *can't figure out how many R's are in \"strawberry\"*\n\nI think we have some time to figure this out 😂
Me: \"AI, write a simple email\"\n\nAI: \"I'd be delighted to craft a professional correspondence that effectively communicates your intended message while maintaining appropriate tone\"\n\nBro, I just wanted \"Hey, can we meet Tuesday?\""
🚨 CONTROVERSIAL: AI tools are making people DUMBER, not smarter
Here's why everyone's getting it backwards:
1/6
5/ The winners understand this:
AI is a tool, not a crutch.
It should accelerate your expertise, not replace it.
The goal is human + AI > human alone, not AI > human.
Most people are aiming for the wrong equation.




