Stepan Zinin

@omooretweets It's very tipical for Moscow too. But robots need to deal with snow and cold weather.

You know you’ve lived in SF too long when this is peak cinema

@elonmusk 🚀 SpaceX launches by year: 2006: 1 2007: 1 2008: 2 2009: 1 2010: 2 2011: 0 2012: 2 2013: 3 2014: 6 2015: 7 2016: 9 2017: 18 2018: 21 2019: 13 2020: 27 2021: 33 2022: 61 2023: 98 2024: 133 2025: 105 as of August 26

@kwindla omg, we literally developed same debug process for our app by ourselves

Are you building realtime AI applications that do complicated things using multiple models and processing stages? (Voice agents, video avatars, copilots, etc.) If so, you've probably done a fair amount of squinting at timestamped log lines to try to figure out what data is flowing where in your computation graph. @aconchillo has been working a little bit on a visual trace log tool for realtime AI and he just made the repo public. Check out Whisker, a @pipecat_ai debugger ...

📊 In other news I also rawdogged an app called error2telegram It sends any errors from PHP, SQLite, Postgres, or system metrics like CPU or disk space getting too high, from ANY of my servers, straight to my Telegram I'm making so many little self-contained apps with AI now, it's really fun

I was part of a team that launched a photo realistic AI image gen mobile app in 2022 and it was crazy how quickly the ROI on ads dropped over a few months as VC-backed companies ramped up their ad spend The way the online ad market is set up these days basically means everyone bids up ad spend until they are breakeven on unit economics and all the profit goes to facebook/google

I already tried ads repeatedly and ROI was always negative I hired the best people in the industry and ROI was still negative I think the reason is basic economics: I compete with VC-funded companies with $100M balance sheets who happily enjoy negative ad ROI if it means growing their user base They can pay $1 per ad, lose $0.10, make $0.90, and it's still good because it means GROWTH. And growth means being able to raise another $100M for the next round. Then at some point they try push to profitability and IPO! I can't do that cause I'm self funded, so everything I do has to be profitable If you're competing for the same ad slot with those companies, you can't economically have positive ROI for very long. For a while yes, not for very long.

It’s pretty amazing that consumer AI hasn’t become “winner take all” (or arguably, even “winner take most”) on the model front I use different models for each modality: - Text = ChatGPT - Image = Flux - Video = Veo 3, Seedance - Sound / voices = Eleven Labs …and even more!

Nobody who’s actually good at making money needs to sell you a course on it.

If you spend 80% of your revenue on ads to make your money Then remove your GPU costs for vision models running on pics of millions of free users Then remove the 30% App Store fee Then have to split what's left with cofounders You might at some point realize selling courses is an easier business

Biggest takeaway from App Mafia debacle is $3.6m MRR doesn’t go very far when you’re renting an audience. Owning your own distribution is more valuable than ever.

Over the past few weeks, we have seen ChatGPT, Midjourney, DALL-E 🤖 and various other AI tools bring people's ideas into reality, including those without real experience. 🌟 We created our first-ever game "Aliya's Ascent" 🎮, using a plethora of AI tools including GPT-4, Midjourney, DALL-E, and more. 🛠️ Here are the prompts and the process we used:

Video game entirely generated by AI, and, in Studio Ghibli style.

Vibe coding is here to stay. I'd been worried it might be a fad, but I talked to the founder of an infrastructure company who's in a position to see how well vibe-coded apps are doing, and he said a lot of them are making money.

@SynthflowAI plz show all 30

🌍 Synthflow now supports 30+ languages — Spanish, Portuguese, Japanese, Korean, Turkish, Vietnamese & more — to connect with customers worldwide. #VoiceAI #MultilingualAI #AIForBusiness

Get a normal job Then make side projects after work Once the side projects make more than your main job for 6-12 months, quit yoir main job I don't know anyone who just fucked around being unemployed and then suddenly built a successful business, it was always people who had a job/income already and then what they built on the side took off On the contrary, there's something about being unemployed, sitting at home, getting free money from government or parents that gives absolutely NO sense of urgency because you don't hate your life enough You really hate your life doing a regular shitty job though, and that's GREAT motivation to work hard on the side on projects to be able to quit and change your life around

Yennefer — Voice AI RPG platform my hobby project - started upgrading Yennefer to a new level. - collected useful repositories - generated the architecture and task list for implementation - started working through the tasks one by one - seems to be going well

Building Voice AI RPG game. Added image generation.

GPT-5 is now available in Cursor. It’s the most intelligent coding model our team has tested. We're launching it for free for the time being. Enjoy!

getting pretty nostalgic for the future

I just talked to a startup whose market cap, if they succeed, will be so large that we didn't even try to estimate it. It would just sound like fiction to investors.