Full-stack developer building SaaS products 🚀 Sharing what I build, learn & discover. Building in public ↓

This is honestly pretty crazy. Solving a nine-loop physics problem mostly on its own for a few thousand dollars is a wild jump.
New on the Science Blog: Yes, Claude can do Nine Loops. Theoretical physicists predict how particles behave using formulas called scattering amplitudes. These are notoriously hard to compute, so researchers work with layers of increasingly fine corrections called “loops”—each added loop makes the answer more precise but takes exponentially more computation. Most calculations stop at two or three loops. Eight loops was the previous record in a simplified model physicists use as a testing ground (planar N=4 super-Yang-Mills), set by SLAC's Lance Dixon and collaborators. Last month, physicist and science writer @4gravitons issued a challenge: could an AI push past eight loops in this model, using only the compute budget an academic could reasonably access? Given a single prompt describing the nine-loop problem, Claude ran largely unsupervised for days in Claude Science and solved it using methods developed by Dixon and his colleagues, at a total cost of a few thousand dollars. Dixon independently verified the result, and von Hippel wrote about the experience for our blog. Read more: anthropic.com/research/yes-c…
28
Q: why do people use multiple AI models? A: because every model has different strengths. but switching between 4 tabs is annoying. what if you could ask all of them at once? #SaaS #AI
1
18
what's harder? building a startup or getting people to notice it? building the product is one thing... getting the first 100 users is another game. #Startup #SaaS
40
One thing I wish every beginner developer understood: **Tutorials make you feel like you're learning. Projects show you what you actually know.** Build something. Get stuck. Search. Fix it. Repeat. That's where the real learning happens.
1
1
33
You don't need to train your own AI model to build an AI product. You can start with: → an existing model → an API → your own logic → a real problem The goal isn't to build the biggest model. **The goal is to build something people actually want.**
21
Okay, a Tamagotchi-style AI on a keychain is actually kinda wild 😂 Curious to see what people end up doing with it.
JUST IN: Meta unveils Muse Charm, a Tamagotchi-style AI gadget designed to fit on a keychain, with Zuck calling it a “joyful little device.”
15
That’s a pretty crazy shift in just two months. Anthropic still being #1 in spend while OpenAI is catching up that fast is interesting.
Spend of OpenAI vs Anthropic vs Open (Vercel AI Gateway, last 2 months) • Anthropic still #1 in spend, but went 69% → 40% • OpenAI: 10% → 24% in spend • GPT-6 Astra + GPT 5.6 Sol are ripping • OpenAI now leads in tokens # • Kimi K3 + DeepSeek took ~half of Anthropic's loss • Opus 5.5 is up to 10% of spend in 2 days • OpenAI is 62% of image generations Watch here: token-race.vercel.app
21
That’s actually fascinating. Finding a new programmable DNA system in bacteriophages could open up some really interesting biology if they figure out what it does. The CRISPR comparison makes this especially worth watching.
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR. We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use. Read more: anthropic.com/news/claude-di…
21
😂 The name definitely grabs attention. Curious to see how well the AI actually handles the roster management side.
Say hi to Sidehoe The first AI assistant to manage your roster Try it now: sidehoe.chat
Community note
This is a parody ad for an app launch video service and is not real per the creator sidehoe.chat/chat/
1
34
That’s pretty wild An AI continuously improving an open-world game while it’s running is a crazy demo of what agentic coding can do.
Opus 5.5 has been looping on my open-world NYC game for almost a day now, and it's continuing to improve. Jump in and try it (it's multiplayer!): somethingbig.ai/world/
72
Q: if you could launch one SaaS today... what problem would you solve? drop your idea below 👇 I want to see what people are building.
36
That’s the crazy part The gap between “Hollywood production” and “one prompt” is getting smaller really fast. 🔥
Hollywood has spent years perfecting scenes like this AI is doing it from a prompt. 💀🔥
26
The interesting part is that they’re basically admitting the current system is too binary. They want to move toward age/region-based controls instead of just making materials private, which seems like a much more practical approach.
現在、CLIP STUDIO ASSETSでは、一部の素材を非公開にしています。運用方針にそって確認を行う中で、非公開となる素材が増えており、ご不便をおかけしております。また、ご説明が不足していることもあり皆さまにご心配やご不安をおかけしております。大変申し訳ございません。 ・管理者による削除や未公開状態に変更となった場合でも、違反素材として削除されない限り、権利関係などの問題はありません。すでにダウンロード済の素材は引き続きご利用いただけます。 ・非公開となった素材は、判定に誤りがあったもの、問題のない状態に修正いただいたものから確認を進め、再公開の予定です。素材作者様ならびに素材をご利用の皆様には大変お手数をおかけいたしまして、申し訳ございません。 ・購入などで素材を取得済の方は、非公開となった素材でも、CLIP STUDIO ASSETSのマイダウンロードから再ダウンロードできるようにする改修を行います。ご不便をおかけし申し訳ございませんが、いましばらくお待ちください。 CLIP STUDIO ASSETSは、現在、多くの国や地域の方に利用されています。利用する方の年齢や文化、地域によって、受け止め方が大きく異なる表現もあります。そのため、センシティブな表現については、単に「掲載できる・できない」だけではなく、誰に、どのように表示するかまで考えることが、これまで以上に求められています。 今回の非公開は、作品や作者の方の表現そのものを否定するものではありません。また、その作品を「不適切な表現」と評価することを目的としたものでもありません。あくまで、CLIP STUDIO ASSETSという公開サービスの中で、現在どのような形で掲載できるかという観点から判断しています。 掲載の判断について、「どの表現が対象になるのか」「どのように修正すれば公開できるのか分かりにくい」というご意見があることも認識しています。掲載の可否は、特定の部位や表現だけで一律に決まるものではなく、素材全体の内容や見せ方、掲載情報などを含めて確認しています。また、掲載の判断で考慮する点や必要な対応は、現在も変わりつつあります。今後、掲載ルールの説明や代表的な例を充実させ、投稿前に判断しやすい情報を増やしてまいります。 本来は、年齢制限や地域ごとの表示制御、コンテンツフィルターなどを使い、より細かくゾーニングできることが望ましいと考えています。今後は、年齢や地域、利用者の設定に応じた表示の仕組みを強化していく予定です。 「すべての利用者に表示する」か「まったく公開できない」か、という二択ではなく、表現の内容に合った形で公開できることを目指します。弊社は、これからも幅広い創作表現を支え、安心して利用できることと、多様な表現を扱えることの両方を目指して、ASSETSの仕組みと運用を改善してまいります。
60
That’s actually wild. 140 years of self-play just to make a robot good at football 😂
We trained a robot to play football. How? Self-play for 140 years in a virtual World Cup. Meet the #Messinator:
40
Q: what's more important for a new SaaS? A: 100 users who love the product or 10,000 users who barely use it? I'd rather have the 100.
1
24
Q: what's the hardest part of building a SaaS? A: building the product is only half the battle. getting users is the real challenge. agree?
2
1
30
50% lower API pricing is a pretty big deal 👀 Faster + cheaper models could make a lot more AI products viable at scale.
Please welcome GPT-6 Sol and GPT-6 Luna to the GPT-6 universe. GPT-6 Sol and Luna build on the advances behind GPT-6 Astra, bringing much of its strengths into faster and more affordable models to support work at scale. We’ve also made caching and inference more efficient, and we’re passing the savings directly to you: 50% lower API prices for Sol and Luna compared with GPT‑5.6 promotional pricing.
22
Q: why do developers use multiple AI models? A: because no single model is perfect. one might be better at coding, another at reasoning, another at writing. so why choose only one?
22
This looks really clean 👀 The no-ads experience is a nice touch. Curious to see how Kino evolves from here.
introducing Kino (again). watch any movie, series or anime for free with no ads. in HD. kino.prince-the.dev
2
60
That’s a serious expansion 😂 At this point, CodexBar sounds more like an AI developer toolbox than just a Codex app.
CodexBar learned a few new tricks in the last few weeks! TypeSafe, Nous Portal, Muse Code, CodeRabbit, Replicate, HuggingFace, Pi, Helmcode, v0, Charm Hyper, GitKraken AI, Bifrost My ambitions for this app REALLY outgrew the app name. (Oh and these keychain alerts are no more) codex.bar
43