🅺evin 🅶oldsmith 🇺🇦

17.9K posts

🅺evin 🅶oldsmith 🇺🇦 banner
🅺evin 🅶oldsmith 🇺🇦

🅺evin 🅶oldsmith 🇺🇦

@KevinGoldsmith

CTO @ DistroKid, Principal @ Nimble Autonomy. Board Member & Speaker. Author of "It Depends: Writing on Technology Leadership 2012-2022." Let's talk on BlueSky.

Seattle Katılım Mart 2008
1.5K Takip Edilen3.9K Takipçiler
🅺evin 🅶oldsmith 🇺🇦
🅺evin 🅶oldsmith 🇺🇦@KevinGoldsmith·
New ep of It Depends: "Every Organization Is a System. Are You Designing It or Just Living in It?" I've been telling people to learn systems thinking for years without ever teaching how. This episode is my attempt to fix that. itdependspod.com/episodes/every…
English
0
1
0
57
🅺evin 🅶oldsmith 🇺🇦
🅺evin 🅶oldsmith 🇺🇦@KevinGoldsmith·
New ep of It Depends: "Talking to Executives: That's Not a Derailment, That's the Meeting." I once got permanently banned from presenting to a VP because I insisted I was right about the word "avatar." I was right. It didn't matter. itdependspod.com/episodes/talki…
English
2
0
0
82
🅺evin 🅶oldsmith 🇺🇦
🅺evin 🅶oldsmith 🇺🇦@KevinGoldsmith·
The adversarial approach is a long-established pattern in machine learning to produce better results. I’m leaning a lot more on it to tame the quantum-like nature of LLM outputs.
zak.eth@0xzak

Just shipped adversarial-spec, a Claude Code plugin for writing better product specs. The problem: You write a PRD or tech spec, maybe have Claude review it, and ship it. But one model reviewing a doc will miss things. It'll gloss over gaps, accept vague requirements, and let edge cases slide. The fix: Make multiple LLMs argue about it. adversarial-spec sends your document to GPT, Gemini, Grok, or any combination of models you want. They critique it in parallel. Then Claude synthesizes the feedback, adds its own critique, and revises. This loops until every model agrees the spec is solid. What actually happens in practice: requirements that seemed clear get challenged. Missing error handling gets flagged. Security gaps surface. Scope creep gets caught. One model says "what about X?" and another says "the API contract is incomplete" and Claude adds "you haven't defined what happens when Y fails." By the time all models agree, your spec has survived adversarial review from multiple perspectives. Features: - Interview mode: optional deep-dive Q&A before drafting to capture requirements upfront - Early agreement checks: if a model agrees too fast, it gets pressed to prove it actually read the doc - User review period: after consensus, you can request changes or run another cycle - PRD to tech spec flow: finish a PRD, then continue straight into a technical spec based on it - Telegram integration: get notified on your phone, inject feedback from anywhere Works with OpenAI, Google, xAI, Mistral, Groq, Deepseek. Leveraging more models results in stricter convergence. If you're building something and writing specs anyway, this makes them better. Check it out and let me know what you think! github.com/zscole/adversa…

English
2
0
2
208
🅺evin 🅶oldsmith 🇺🇦
🅺evin 🅶oldsmith 🇺🇦@KevinGoldsmith·
New podcast episode for the new year! Why a reflection practice is critical as you move into senior leadership (because no one else is going to be focusing on developing you). You can’t lead others to grow if you’ve stopped growing yourself. pod.link/1724860326/epi…
English
0
0
2
46
🅺evin 🅶oldsmith 🇺🇦
🅺evin 🅶oldsmith 🇺🇦@KevinGoldsmith·
Tech debt isn’t a moral failure. It’s a signal Some debt is deliberate Some is harmless Some actively constrain the business Some is code telling you a people story If you treat it all the same, you’ll make the wrong calls If you read the signal, you won’t kevingoldsmith.substack.com/p/tech-debt-is…
English
0
0
0
48
🅺evin 🅶oldsmith 🇺🇦
🅺evin 🅶oldsmith 🇺🇦@KevinGoldsmith·
@jnack @claudeai @NanoBanana I’ve had both very good and very bad luck generating google slide decks with Gemini. Sometimes, it has turned an outline into a usable, editable, deck. Sometimes it gives me slide-by-slide instructions on how to make the deck myself or tried to write code to generate the file.
English
1
0
0
57
John Nack
John Nack@jnack·
.@claudeai just made me a super crappy deck. Silver lining: I discovered the new Google Slides ability "Beautify slide" via @NanoBanana 🍌🪄🙏:
John Nack tweet media
English
2
0
3
387