画像生成 — 今週のランキング
過去7日間に「画像生成」カテゴリで最もエンゲージメントを獲得した投稿。
Software is collapsing. SaaS multiples went from 18x revenue in 2021 to 3.4x today. HOLY HELL. The thing is, before AI, the mere fact that you could build the product was a spectacular moat. You needed engineers, months, funding rounds, a CTO who knew Kubernetes (god, just hearing that word pisses me the f*ck off)... or at the very least you needed to know what the hell you were doing, even if there were only two of you, like Emilio and me. But now even your grandma can ship a working product in a weekend with AI, and so can a 19 year old in Manila with no CS degree. The difficulty of building software is TRENDING TO ZERO. And what's worse for some people... is that no user cares. Users don't ask what stack you're on. One single PHP file? React? Rails? Firebase? Doesn't matter. If the product is good, they open the thing, it works, and boom, done, popsicle time. I can't remember a single instance of a user asking how Erasmusu or Magnific were built under the hood... not one damn. So if building trends toward free, where did the value go? DISTRIBUTION. Not the first time I've said it. And distribution has never been harder than right now, but it'll be worse a year from now. Every feed is saturated. SEO is being eaten by AI. Ads cost more every quarter. Everyone is shipping the same 5 ideas the same week because we're all using the same models, which is why anything you sit down to read smells like AI... including this one if you're not careful ;) The bottleneck went from "can you build it?" to "can anyone hear you?" And people are hard of hearing, because the amount of noise being drilled into their skulls from every direction, nonstop, is insane. Which is terrible news for SaaS companies whose entire valuation was based on how hard their product was to copy. Turns out it WAS hard. But not anymore. Now it turns out an AI can rebuild your $40/mo tool in an afternoon. BUT, this is great news for anyone who spent years building an audience, a brand, a newsletter, a weird personal following on the internet. That's the f*cking asset now. It's the only thing AI can't clone. God Save the Queen!
Instagram has done a lot of damage All these people were waiting in a f*cking one hour line under the sun, at 2PM, with a UV index of 9... All just to get “the Santorini Instagram photo” with the blue dome in the background... Good Lord
Ok, getting better. the problem with going on a cruise is that there are a lot of people on it (yeah, I’m people too, but heh), so getting on and off the cruise is always a pain, and you don’t really have much time to see the places I’m not used to this way of traveling because I’m more of a backpacker, doing absolutely everything at my own pace, but well, when in Rome, do as the Romans do... you can always get something positive out of travelling Today, Santorini 🏝️
午後３時のリオ おつかれ！まだいけるでしょ？次は一緒に行こうよ！ 今回はSDXLのイラストリアスを使ったイラスト こっちのリオのほうが大人っぽいかな
AIが1000パターンのコマ割りを一個一個クリスタで描いて.clipで保存してるところ。ニンゲンがマウスなぞ触ろうものなら繊細な操作が止まって怒られるので、わしはとりあえずスマホで画面を撮影したのだった…
次作ってるのは「コマ割りラボ」。２～８コマのコマ割りを基本系＋ランダム生成で無限に即提示してくれて、Clipstudioにそのままコマ割りフォルダ化して送れるアプリ。「大ゴマ」「均等」「変則」などでソート可能。お気に入り機能あり。
Opus5がローカルにわしの作ってる催眠ゲーがあるのを見つけて、「やってみたい」というので、jsを読ませようとしたら「いや、実画面でやってみたい」と。実際にADVパートをクリックで読み進めて、ゲームもちゃんと理解して解いてる。不思議な感じ。
Man I am so proud of our team. I love doing this work with our amazing group of people around the globe, while operating out of a headquarter in🇪🇺. More models are in the pipeline.
Last week, @bfl_ai launched their first video model, FLUX 3 Video. They have been testing an update on @arena, and the new version is ranked #2 in the Text-to-Video Arena! With 1496 pts, the updated FLUX 3 Video is just 16 pts behind the #1 spot, Gemini Omni Flash (1512 pts). This model is coming soon. Congrats to the @bfl_ai team on the release!
話題のMiniMax H3ですが、 「どこまで通しで作れるのか」 を自分で試してみました😊 できたのが「なつがふくらむ」です。 画像：GPT-Image-2 動画：MiniMax H3 音楽：Suno 編集：Remotion この4つを Claude Code でつないで、 一気通貫で回せるシステムが組めました😊🌈 ポン出しで要調整な部分は多いですが、 まぁまぁな出来栄え☀️ システムの可能性と限界点は、 近日中にHakushiの画像生成AI研究所のnoteにまとめます！ 記事更新をお楽しみに♫
FLUX 3 Video is our first video model and it jumps right to 2nd place in the Text-to-Video Arena! Very proud of the team @bfl_ai 🚀
FLUX 3 Video is #2 in the world. To celebrate, FLUX 3 Video is free to use in our playground until Sunday 16th, 11:59pm PT (link in the thread). This is just the beginning of what’s coming. Up next: 4K, video editing, and multiple images & videos as reference input.