IKcode | Igor Wnek

"If your team is too busy doing their 'normal job' to experiment with AI, you're preparing them to be replaced." Geoffrey Huntley wrote that last week. I agree, and I'd go further.

Then show your work. Publish the repo, write a short post, share what broke. One piece of advice: this week, build a small proof of concept you've always wanted, with a coding agent. Then publish it, together with what you learned.

"If your team is too busy doing their 'normal job' to experiment with AI, you're preparing them to be replaced." Geoffrey Huntley wrote that last week. I agree, and I'd go further.

Using AI at work isn't enough. Build the thing you've always wanted to build. With coding agents, that takes evenings, not months, and that experience is what makes you valuable at work.

"If your team is too busy doing their 'normal job' to experiment with AI, you're preparing them to be replaced." Geoffrey Huntley wrote that last week. I agree, and I'd go further.

I still know engineers who say they won't use AI. AI won't replace them. Engineers who use it every day will. Without AI, the work is too slow compared to theirs. They try more ideas, ship more, and still have time to check what they ship.

Claude Haiku 5.5 is out, almost a year after Haiku 4.5. On paper, it was worth the wait. Per million tokens, for prompts up to 100K: Input: $0.10 (Haiku 4.5: $1) Output: $0.50 (Haiku 4.5: $5) Longer prompts cost half as much as on Haiku 4.5. Anthropic puts the average saving at around 75%.

Same update: Sonnet 5.5 cache reads drop from $0.20 to $0.10 per million tokens. Anthropic puts the saving at around 20% on most agentic work, in line with my own estimate. With Haiku taking the right subagent work, I expect my Claude Max plans to go much further.

Claude Haiku 5.5 is out, almost a year after Haiku 4.5. On paper, it was worth the wait. Per million tokens, for prompts up to 100K: Input: $0.10 (Haiku 4.5: $1) Output: $0.50 (Haiku 4.5: $5) Longer prompts cost half as much as on Haiku 4.5. Anthropic puts the average saving at around 75%.

Could Gemini 4 Argon replace Opus 5.5, Sonnet 5.5, GPT-6 Astra, or GPT-6.1 Sol? Access is still restricted. Argon leads Google's DeepSWE comparison, but trails Opus and Astra on FrontierSWE. I'd evaluate specific tasks before changing a default.

Vals' Mini-SWE-agent terminal runs put Argon behind Opus, Sonnet, and Astra. It scores above Sol 6.1, but costs about 10x as much per test at the published API rates. Some Claude tasks used fallbacks. That doesn't establish value under my subscriptions.

Re-reading "The 7 Habits of Highly Effective People". The 1st Habit - "Be Proactive" - reads differently when you're a product engineer, an AI engineer, or principal-ish on your team. It's not "take initiative". It's about where you choose to spend your attention.

Covey's distinction: Circle of Concern (everything you care about) vs Circle of Influence (what you can actually affect). Most engineers I know - drown in the Concern circle. Model news, org politics, every team's stack choices.

Copilot via VS Code or Copilot CLI? What's your choice? #copilot #vscode

Recently I've been experimenting a lot 👨‍💻 with different ways of creating customisations for Copilot, HeyAlice app, and interactions with LLMs in general (system prompts, prompts). The final conclusion 🎯 is - XML-like tags are 💯 clear winner. #copilot #vscode #context #anthropic #openai #ai

Recently I've been experimenting a lot 👨‍💻 with different ways of creating customisations for Copilot, HeyAlice app, and interactions with LLMs in general (system prompts, prompts). The final conclusion 🎯 is - XML-like tags are 💯 clear winner. #copilot #vscode #context #anthropic #openai #ai

If you are into #copilot and #vscode, you might not know that I work on my copilot-kit repository. It's just been updated with really nice prompt to generate single custom instruction, works nicely for both greenfield and legacy apps. 1/2

Busy half of the Sunday is almost done! 👨‍💻🚀 As usual I was working on some new ideas and improvements to my workflow with #vscode and #copilot. The first public deliverable is the new #backendengineer agent for #copilot: - github.com/ikcode-dev/c...

Busy half of the Sunday is almost done! 👨‍💻🚀 As usual I was working on some new ideas and improvements to my workflow with #vscode and #copilot. The first public deliverable is the new #backendengineer agent for #copilot: - github.com/ikcode-dev/c...

Yesterday we've posted new video covering #vscode #copilot and the new model we can FINALLY use in the #agentmode - Claude Opus 4.5 #claudeopus Previously Opus models were available only in the ask mode - not that useful, at least to me. So it's kinda revolutionary update! 🚀 youtu.be/mP9WDVYdUnc

Ha! I'm amazed how well Claude Opus 4.5 works in #vscode #copilot. 🤩 Remember to try it out because there is a promo right now - Claude Opus 4.5 costs only 1 premium request in #copilot. What is better, for few days Opus 4.5 is finally available in the #copilot #agentmode - fantastic news! ❤️

I have so many different ideas. 🧠 Before I used #linear to handle all my important #projects, I often ended with working on something one day, to drop the idea completely the next day. Then I started working on the next idea, which I quickly abandoned again. Now I can #focus and DELIVER. 🚀

Hello Team! Long time no... write? If you are into #copilot and #vscode - I added new prompts to my repository - github.com/ikcode-dev/c... I use them at day-to-day job to improve code quality and they work great. Feel free to test them out! Let me know what you think. 🤔