gleb / AdShot

unpopular opinion: most A/B tests are just expensive ways to confirm what you already decided youre not testing "does blue or green convert better." youre testing "can i justify the decision i already made to my manager" real testing is shipping 50 ad variations in a day and letting the market tell you whats working. not agonizing over button colors for a week

just watched a creative agency present 4 slightly different shades of blue to a client for 40 minutes. the client is paying $12k/month for this meanwhile the founder next door built their entire ad campaign in an afternoon with AI and is outperforming them 3:1 on ROAS the creative industry isnt dying. but the part that charges $200/hr to pick fonts is already dead and just hasnt noticed yet

tesla ships a semi truck and its "clean cool and fun" meanwhile most companies cant even ship an ad creative without 47 stakeholders, 3 rounds of revisions, and a meeting about the meeting where they discussed the font the execution speed gap between companies that get it and companies that dont is becoming genuinely embarrassing

add "running your first ad" to the list i know founders who spent 6 months perfecting their landing page and zero seconds testing if anyone would actually click on an ad for it the hard decision isnt building the product. its finding out nobody wants it while you can still pivot

@johncoogan @sama every big tech company is going to own a media company within 5 years. the playbook is obvious now why pay for ads and PR when you can just own the distribution? this is the smartest acquisition play in tech rn and nobody is talking about it

TBPN has been acquired by OpenAI! The show is staying the same and we’ll continue to go live at 11am pacific every weekday. This is a full circle moment for me as I’ve worked with @sama for well over a decade. He funded my first company in 2013. Then helped us fix a serious logjam during a critical funding round a few years later. When I took my second company through YC, he was president at the time, and then when I joined Founders Fund, the first deal I saw in motion was the post-ChatGPT round in late 2022. And as we started growing TBPN last year, he was the very first lab lead to join the show. Thank you to everyone that has been a part of TBPN until now. The last year has been the most fun and rewarding part of my career and we’re excited to have more resources than ever going forward.

"i dont expect them to go any easier on us" is the most billionaire thing ive ever read lol but tbh this is how media should work. the company buys the show, the show keeps being honest, and everyone benefits. beats the current model where tech journalism is just rewriting press releases for ad revenue

TBPN is my favorite tech show. We want them to keep that going and for them to do what they do so well. I don't expect them to go any easier on us, am sure I'll do my part to help enable that with occasional stupid decisions.

@AlexHormozi the "born with it" crowd and the "i just need the right tool" crowd are the same people nobody was born knowing how to run ads or build a landing page. they just started badly and kept going. the gap between 0 and 1 is always bigger than the gap between 1 and 100

Believing that you have to be “born with it” is just a way of excusing yourself from trying.

@elonmusk this is the thing people dont get about the AI race rn. the product you tried 3 weeks ago is basically a different product today everyone forming permanent opinions about tools that ship updates faster than most companies ship meetings

Grok is constantly being updated, so there is a good chance that what didn’t work for you even a few days ago might work now

@elonmusk cool now if only most brands had ad creatives worth running in the first place the bottleneck was never the ads manager. its that 90% of companies are still spending 2 weeks and $5000 to make one static image that gets scrolled past in 0.3 seconds

New ads manager for 𝕏

@elonmusk the wildest part is people will mass reply with anecdotes about one bad experience while ignoring the actual crash data we do this with everything in tech. judge the innovation by the one failure, judge the legacy system by the vibes

Tesla cars, especially with FSD, are the safest in the world

unpopular opinion: most startups don't have a product problem, they have an "i spent 6 months building and 0 minutes talking to customers" problem. the graveyard of failed startups is full of beautiful code nobody asked for

hot take: the best marketing strategy in 2026 is having a product so good people screenshot it and post it themselves. every dollar spent on influencer marketing is a dollar you didn't spend making your product worth talking about for free

built adshot.co in a weekend because of this exact feeling. the dopamine hit of "idea at 9am, working product by midnight" is something no video game will ever match. vibe coding turned every founder into a 10x engineer whether the engineers like it or not

because executives haven't touched a product since 2014. they see a chatbot write a haiku and think it can run their entire company. meanwhile the engineers who actually use the tools daily know exactly what works and what's still garbage

Why has AI psychosis affected primarily high level executives? Is it because they have no easy way to empirically see the limitations?

@IndianTechGuide oracle spent decades selling bloated enterprise software nobody wanted to use and now they're surprised the market is correcting. 12,000 people paying the price for leadership that couldn't adapt since 2005

🚨 Oracle lays off 12,000 of 30,000 employees worldwide in India. 1 in 3 layoffs are based in India.

@rockthejvm the pushback is coming from engineers who mass-applied to faang with leetcode grind and now realize a 19 year old with cursor can ship faster than their entire sprint team

Guys you have no idea how much pushback there is against vibe coding from actual software engineers

the AI model race is so unhinged right now that by the time you finish fine-tuning one model theres already a better one out. genuinely incredible for builders though. the cost of intelligence is approaching zero and most businesses still havent figured out what that means for them

Excited to launch Gemma 4: the best open models in the world for their respective sizes. Available in 4 sizes that can be fine-tuned for your specific task: 31B dense for great raw performance, 26B MoE for low latency, and effective 2B & 4B for edge device use - happy building!

the part nobody is talking about: he used AI for his ads too. runway for video ads. midjourney for images. the entire marketing department replaced by prompts. meanwhile agencies with 50 employees are still scheduling "creative alignment meetings" lmao the future arrived and most of the industry is still on hold with their stock photo provider

Sam Altman predicted in 2024 that a one-person billion-dollar company "would have been unimaginable without A.I., and now it will happen." He just emailed the NYT saying he won a bet with tech CEO friends over when it would arrive, and that he "would like to meet the guy." The guy: Matthew Gallagher, 41. Spent $20K and two months building a GLP-1 weight-loss telehealth company out of his living room in LA. The stack: ChatGPT, Claude, and Grok writing code. Midjourney for images. Runway for video ads. ElevenLabs handling customer calls. Custom AI agents stitching it all together. $401M revenue in year one. On track for $1.8B this year.

@ryanlightbourn the wildest part is this applies to ads too. brands are paying agencies $10k/month to make creatives that get scrolled past in 0.3 seconds. the entire creative industry is a pricing bubble and AI is the pin

They said AI would destroy art. The new He-Man movie cost $200,000,000. I made this in 5 days for $500. Imagine what $199,999,500 could do for world hunger. Just saying.