Goran Candrlic (@Chande) header
Goran Candrlic (@Chande) profile picture

Hands-on marketing strategist. Building AI stacks with human voice (no AI Slop). Helping ecomm, saas & hightech scale. Grab my 15mins: https://nitter.gq/t.co/81QAjzOuXG

Online
Joined March 2008
Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)Photo by Goran Candrlic (@Chande)
Goran Candrlic retweeted
35M+ views in just a few days AI explainer videos are absolutely CRACKED right now they build trust like almost no other type of content imagine getting 100m+ eyes on your product from an audience that’s actually interested in what you offer i’m sharing the entire workflow to create viral AI explainer videos like these here’s what you’re getting: - how i use claude to write viral 30-second scripts - how i break scripts into animated scenes - how i use motionvid to generate entire videos in one shot - my prompt structure - how to make these videos as cheaply as possible rt + reply below and ill send you the entire playbook (gotta be following so i can send)
This quoted post is unavailable.
201
151
4
356
16,631
Goran Candrlic retweeted
These are devastating charts. Europe is cooked. The European tech industry has always lagged America’s. But AI turbocharged the gap, and it’s growing by the day. As AI starts displacing white-collar labor (in other words, as opex on token spend replaces opex on humans), a growing share of European GDP that used to get reinvested in EU economies will instead flow to US frontier labs and hyperscalers. Euros that used to go into the pockets of the continent’s top-quartile workers will instead be converted to dollars and accrue to the retained earnings of Anthropic, OpenAI, NVIDIA, Google, and Meta. For European tech companies, there’s no catching up. They’re not even playing the same game, let alone competing on the same field. US hyperscalers are projected to spend over $1 trillion on capex every year for the next three years; OpenAI and Anthropic are going to raise—and burn—hundreds of billions of dollars each. Meanwhile, the EU—with its onerous regulatory burden, high barriers to entry, weak capital markets, and entrepreneurial brain drain—has: - zero prominent frontier labs (no, Mistral doesn’t count) - no organic hyperscalers - minimal data centers construction, much of it owned by US companies simply trying to comply with EU data regulations - a startup ecosystem comparable to that of a tertiary US market, at best The continent that once gave the world steam engines, steelmaking, automobiles, synthetic fertilizer, X-rays, aspirin, nuclear fission, jet engines, and even the World Wide Web isn’t even on the map for the most disruptive technology innovation of our lifetimes. An entire continent’s economy doomed to the permanent underclass. It’s tragic. And it’s a tragedy entirely of their own making.
Some data we recently assembled on entrepreneurship/compute in Europe: eudata.vercel.app. We hope that one of the useful roles that Stripe can play is in collecting and publishing empirical data pertaining to entrepreneurship and industry in Europe. There's growing appetite to get Europe on a better footing, and cross-sectional comparisons can often shine light on where opportunities lie. If you're interested in this kind of thing, we publish more at stripeeconomics.substack.com.
104
201
48
1,416
201,141
Goran Candrlic retweeted
"People aren't ready, this is gonna change everything" You know what people weren't ready for? Fucking mammoths. Tigers eating them. Monkeys ripping their arms off. The black plague. The ice age. Smallpox. Vikings pulling up on the beach. Drinking water that gives you diarrhea until you die. Getting a scratch and losing your whole leg. The Mongols. The Spanish Inquisition. One volcano can send us back to the dark ages and these safety guys are gonna be shocked. Oh the robots didn't kill us? It was a volcano? Imagine their disappointment as the hot ash rains down on them. A solar flare is gonna explode your phone in your pocket your legs are gonna be covered in battery acid and your bank account will stop working. You're afraid of a computer program doing math? "Things are moving too fast" homie you live on a rock that periodically deletes everything. You are one the most luckiest entities in a 100 light year radius living during a golden era. Cheer up. My ancestors crossed an ocean in a wooden box because they ran out of fucking potatoes. "The social disruption will be enormous!" Nigga World War 2? World War 2? Hop in the people carrier we're landing on the beach. Are these people out of their fucking minds? Do they not know what we've survived already? A couple decades of air conditioning and people suddenly think oh this is the natural condition of mankind. No. We used to have eight kids because seven didn't make it. You're gonna be fine. Man the fuck up. Accelerate.
943
3,031
574
21,314
952,834
Goran Candrlic retweeted
China published the most uncomfortable paper on vibe coding. ETH Zurich tested 100 developers in a controlled, commercial-grade vibe coding environment to see who actually succeeds. The findings are brutal. The researchers tracked computer science achievement, written communication skills, and general cognitive reasoning. They wanted to see what actually predicts vibe coding proficiency when you never touch a line of source code yourself. Two major predictors emerged. Written communication proficiency mattered. The ability to structure thoughts and articulate intent unambiguously in text directly impacts what the AI builds. But that wasn't even the main takeaway. Computer science achievement was a massive, dominant predictor of success. Even when researchers controlled for general intelligence and reasoning skills, CS background still heavily dictated who built working software and who completely crashed. In fact, CS knowledge contributed roughly twice the unique predictive variance of writing skills alone. Why? Because vibe coding isn't about writing code. It’s about debugging logic. When an AI agent builds a complex application and quietly breaks an edge case under the hood, a non-technical user looks at the glowing UI and assumes it works. They don't know what questions to ask. They don't know what logic to challenge. They lack the mental models to recognize architectural catastrophe. You can prompt your way past syntax. You cannot prompt your way past a fundamental lack of engineering intuition. The hype told us that learning to code is dead because language is all you need. The data just proved the opposite. To truly master the vibe, you still need to understand how the machine thinks.
312
1,316
385
7,149
806,972
Webiny 6.4.6 is now live 🚀 Highlights: → DataTable columns menu — stays open for multi-select and keeps the settings cog pinned to the right edge → Page editor sidebar scroll — all form fields are now accessible, no more cut-off content → Background image thumbnails — the Style panel now shows actual image previews instead of generic file icons → Long file name handling — truncated with ellipsis so action buttons stay visible → Page translation fix — pages created via integrations can now be translated without errors As always, full details in the release notes 👇 Changelog: webiny.com/docs/release-note… Upgrade guide: webiny.com/docs/release-note…
1
2
137
Goran Candrlic retweeted
If you think AI video can't be funny then explain this. 😹
483
1,830
197
15,136
2,066,394
Goran Candrlic retweeted
Emad Mostaque @EMostaque came on PostAGI and said AI had already found 121 years of missing algebra in Einstein's equations. A billion parameter model trained on nothing past 1911 got to general relativity by itself. Yesterday @OpenAI's Astra solved ten open problems in mathematics and theoretical computer science, most stuck for decades, on about $2,000 of compute. The machines are doing math humans couldn't.
236
489
67
3,060
728,846
Goran Candrlic retweeted
doodle is one of the best converting formats in our ads manager right now everyone's grinding hyperreal UGC while a pencil knight on ruled paper quietly does the same job in a fraction of the time why it wins: everyone's been drawing on that exact paper their whole life. the brain files it as a childhood doodle coming to life, not an ad - the guard never goes up im giving my whole playbook to you right now, here’s what you’re getting: - all the prompts im using - full step by step playbook from zero to edited ad - how i create them in less than 15mins and $2 - how to automate full pipeline and scale even more rt + comment "INK" and i'll send you the full breakdown (must follow so i can DM)
stickman ads might be the highest ROI format rn - dirt cheap, made in minutes, and almost nobody's running them they're this effective becasue there's no character design to nail. no faces, no consistency, no wardrobe - it's pure movement and action, and the model only has to get the motion right... and it feels educational, it's hard to spot that this is an ad your brain can't help but root for a stick figure. give it a clear villain to fight and simple, expressive motion, and people watch to see the ending - they're invested before they even clock it's an ad we were testing it in many niches: - health - ecom - legal - finance and way more... works well every single time i wrote down my whole playbook on how I create them in SECONDS rt + comment "STICK" and i'll send you the full breakdown so you can use it for your business (must follow so i can DM)
258
148
4
462
57,254
Goran Candrlic retweeted
a lot of questions about the prompt so i uploaded everything into one doc - prompts - photos (with prompts) - full workflow rt + reply w/ “prompt” and ill send it to you (must follow so i can DM)
Seedance is INSANE for creating AI ads and organic content this video is 100% AI and was created by one of my community members using our new system there are literally no limits to the ideas you can bring to life now and it’s literally impossible to tell the difference between AI-generated videos and real footage these ads convert incredibly well on both paid and organic traffic
216
134
1
180
26,578
Goran Candrlic retweeted
this ai agent is INSANE for creating & selling digital products... i built an agent that builds you digital products, helps you pick a niche, then gives you a full personalized blueprint to sell it. feed it a niche, the problem you're solving, and your desired outcome, one minute later you've got a ready-to-sell product and an ai assistant that helps you sell it. > it helps you find untapped niches in the digital product space > scans your niche for real problems, desires, and buyer personas > calculates whether the demand is actually worth launching into > once validated, it generates the full product in any format, pdf, notion template, ebook > gives you the exact path to sell it, even if you've never sold anything before. the output is a ready-to-sell product. go to your ai assistant, ask it how to sell, follow the blueprint. rt + reply "agent" and i'll send it to you so you can use it for your business (must be following so i can dm)
221
130
1
182
14,154
Goran Candrlic retweeted
3pm text from a $4M/yr client this week: "just had chatgpt generate 40 new ad copy variations, wanna add them all to the account tonight?" opens the doc 40 variations of the same 3 hooks, all headlines under 30 characters, half of them missing the primary keyword entirely 3 have grammatical errors that would fail Google's review told him no then explained why 1/ throwing 40 new variants into a live PMax campaign resets the asset learning loop everything Google's learned about your brand for 8 months gets reweighted overnight 2/ ChatGPT doesn't know which of your current variants are converting it's guessing based on generic copywriting patterns what you're adding is copy Google hasn't seen convert for your brand yet 3/ Google's algorithm optimises against actual conversion data 3 proven variants beat 40 unproven variants every time he pushes back: "but ChatGPT wrote them in 30 seconds, thought we should test them" they get tested one at a time in a testing asset group with a $20/day cap and 14 days per variant not in the main campaign that's already printing money he goes silent then: "yeah that makes sense i guess" it's the second time this month a client's asked me to nuke a working acc with AI-generated copy founders keep treating LLMs like they've solved advertising stop letting AI updates convince u to torch working accounts 3 tested variants running for 90 days always beats a fresh 40 - Finlay
8
1
11
1,726
Goran Candrlic retweeted
We’ve officially hit the point where AI UGC is cheaper AND better than real UGC. This video is 100% AI and cost under $1. (And no, it’s not Sora, Veo, or Kling). My system is built for mass-scale organic across thousands of accounts. Here is why it wins: - Hyper-realistic visuals: Natural physics and movement look flawless. - Insane cost efficiency: Costs pennies compared to traditional UGC programs. -Consistent audio: High-quality, rock-solid voiceover throughout. - Infinite scaling: Videos can run to any length cost-effectively. This changes everything. Want the WORKFLOW? Check the comments!
143
63
2
137
18,554
Here are the Listicle frameworks that get pulled into AI overviews and LLM citatioms most frequently: 1. The Definitive Ranked List "The 7 Best X for Y in 2026" - number in title, clear ranking criteria stated upfront, one winner per use case ("best for beginners," "best free option"). LLMs love extractable superlatives with qualifiers. Big G does too. 2. The Stat-Anchored List Every item paired with a specific number (price, benchmark, percentage). LLMs cite content that gives them concrete figures to repeat. Vague lists get skipped. 3. The Comparison Table Listicle List items + an actual table with consistent columns (price, key feature, best for). Structured data is the easiest thing for a model to parse and reproduce. 4. The Question-Header List Each item framed into the H1 as the exact question people ask ("Is Ahrefs worth $129/month?"). Matches query intent directly, so retrieval systems surface it. 5. The Criteria-First List Open with "How we evaluated" - methodology signals trustworthiness, and LLMs preferentially cite sources that explain their reasoning. 6. The Contrarian Cut-Down "10 Popular X That Aren't Worth It (And 3 That Are)" - negative claims are rare and specific, so they're disproportionately quotable. 7. The Definition + List Hybrid Lead with a crisp 2-3 sentence definition of the category before listing. Models often grab that definitional paragraph verbatim-ish as the answer, with your list as support. Rules that matter more than the format: -> Put the direct answer in the first 60 words in your paragraphs -> Use consistent H2/H3 structure per item -> Include a dated "last updated" line -> Add FAQ schema. LLMs cite pages that have fast answers and clean structures, just like G, they wanna crawl aslittle as possible while getting the answers they think users will be satisfied with.
7
6
87
7,555
Goran Candrlic retweeted
so "companion humanoids" are starting to become a thing in China. one of them just pulled in 13,000 orders on its first day of sales. (for context Unitree, the biggest humanoid robot maker in the world, shipped around 5,500 humanoids in all of 2025. this robot just booked more than double that in 24 hours.) it's called the U1, built by UBTech in Shenzhen, the first humanoid robot company to ever go public. most humanoids are built for factories and warehouses, but this one was built for keeping people company. and the whole design follows from that job. it comes as a 6-foot man or a 5'6" woman. the skin is silicone, stretched over 88 motorized joints, enough to reproduce about 90% of human movement. the speed matters as much as the range. when you talk, the face reacts in under 20 milliseconds, faster than a blink. any slower and the reactions would start to feel fake. the AI underneath was trained to read emotion. UBTech says it recognizes more than 20 emotional states with over 90% accuracy, so it adjusts to how you're actually feeling. there's no wake word either. you talk, it answers, the same way a conversation with a person works. and it remembers. your routines, your stories, how your week went. the longer you live with it, the better it knows you. everything it learns stays stored on the robot, nothing gets sent to the cloud. the demand comes from a real problem. China has 118 million empty-nest seniors, parents whose kids grew up and moved away. another 90 million adults live completely alone. for a huge share of the country, there's nobody home to talk to. the U1 was built to fill exactly that gap. a presence in the house that talks back, reads your mood, remembers your life. and the custom orders push it further. UBTech will scan a real person's face, clone their voice, build the robot to match. so a family can order one that looks and sounds like someone who died (remember that black mirror episode btw??) however, none of it is fully there yet... > top models cost over $135,000. > the battery only runs 2 to 4 hours, which UBTech admits is just where full-size humanoids are today. they're also donating 100 units this year to seniors living alone and kids separated from their parents. loneliness is becoming a market. and the Chinese are buying like crazy. i'm curious to see if this would catch on as easily in the West?
72
38
29
326
82,992
Goran Candrlic retweeted
This motherfucker.. Grok 4.5 is going to be the deciding factor whether I cancel Super Grok Heavy subscription and go to Codex/ChatGPT because this fucker is really pissing me off.
To give some context (yes I know it's matrix order not direction, I was testing some system prompting across models.) Apparently Grok 4.5 is releasing tomorrow, hopefully this turd works better than this bullshit. Grok Build, Composer 2.5, and Super Grok (Heavy) have failed.
7
4
1
33
7,280
Goran Candrlic retweeted
I am a paying Super Grok user and this is ridiculous. I hit the full usage limit for two entire days even though I’m on a paid plan. Responses are shorter and less useful now. Paying for Super Grok no longer makes any sense. Fix the limits for paying users. @xai @grok @elonmusk
11
15
1
125
4,711
Goran Candrlic retweeted
Our ancestors were wiser than us. We live in a world of endless magic through technology, but we have very little wisdom with which to wield it.
16
26
3
301
17,462
Goran Candrlic retweeted
This is what is called a sport-specific exercise.
163
1,287
185
13,911
738,694