dandaka

5 ms·

61.
by dandaka 6mo ago
It will come naturally! I have started with autocomplete as well. I was stumbling upon different problems and was fixing them by implementing best practices. Current stack is: 1/ Claude Code with yolo mode 2/ superpowers plugin 3&
62.
by dandaka 6mo ago
autopilot (yolo mode) is amazing and feels great, truly delegate instead of hand-holding on every step
63.
by dandaka 6mo ago
Interesting that you rate Claude Opus 4.6 lower than 4.5 and 4.7, while community consensus puts it on top.
64.
by dandaka 6mo ago
Any research supporting this? And proposing guidelines / best practices? What exactly are we banning and why? What is allowed and has little risk? Asking as a parent.
65.
by dandaka 6mo ago
I guess it depends on a task. Opus is already spawning Sonnet/Haiku for simple tasks with a good success rate.
66.
by dandaka 6mo ago
1/ Reviewing code can't be more intense than writing. I can't understand where this statement comes from! If that would be true, why would senior developers review code of junior, instead of writing themselves from scratch? 2
67.
by dandaka 6mo ago
Writing code is 10-100x faster, doing actual product engineering work is nowhere near that multipliers — no conflict!
68.
by dandaka 6mo ago
Could be a great feature, can't wait to test! Tired of other models (looking at you Opus) constantly stuck mid-task lately.
69.
by dandaka 6mo ago
I can argue that disaster started mid-4.6, when they started juggling with rate limits while hitting uptime problems. Great we have some healthy competition and waiting for the next move from Deepmind.
70.
by dandaka 6mo ago
SOTA models get distilled to open source weights in ~6 months. So paying premium for bleeding edge performance sounds like a fair compensation for enormous capex.
71.
by dandaka 6mo ago
Can I use Infisical cloud vaults with Agent Vault? I like the UI of secret management there. I like that I can manage secrets from many environments in a single place.
72.
by dandaka 6mo ago
no, most likely you will try codex and get more value (as I did)
73.
by dandaka 6mo ago
Waste of effort... of Opus? If "Opus effort" is cheaper, than dev hours managing yourself more dumb/effective model, what is the point?
74.
by dandaka 6mo ago
If model provider believes they have a better model, it can be a viable bet. But many (me included) started experimenting with other providers because of enshittification from Anthropic (price + uptime). Only to find, that Codex is not that
75.
by dandaka 6mo ago
Every three months I'm trying to deploy to Cloudflare from Monorepo and I hadn't have success yet. While Vercel works every time from the box. Maybe I could dig deeper and try to understand how it works, but I'm super lazy to
76.
by dandaka 6mo ago
I have started to notice this malware paranoia in 4.6, Boris was surprised to hear that in comments, probably a bug
77.
by dandaka 6mo ago
API tokens are complete mess on CF. There are 2 or 3 types of them. Agents constantly confuse, which one does what. Documentation is not referring to them correctly. I still don't understand the difference myself and can't explain
78.
by dandaka 6mo ago
I'm building Kardo ( https://kardo.cards/ ), AI tutor that listens to your calls with a teacher and creates cards for repetition. I live in Lisbon and I've been learning Portuguese with a tutor since 2022. After eve
79.
by dandaka 6mo ago
I am super bullish on Google, they are my best bet to earn from models. Mostly because they are vertically integrated (other revenue streams) + open to provide services to other companies (Apple deal).
80.
Ask HN: Local-first meetings recorder and transcriber?
7 points by dandaka 6mo ago|1 comments
81.
Anti-distill tool, digitize your colleague before they digitize you(twitter.com)
5 points by dandaka 6mo ago|0 comments
82.
by dandaka 6mo ago
You are bot!
83.
by dandaka 7mo ago
Ask your agents, come on!
84.
by dandaka 7mo ago
Still growing YoY even in the most saturated markets (Spain).
85.
by dandaka 7mo ago
Do you play padel in Sweden? There was a big oversupply of courts, powered by bad expectations and venture capital. Things have scaled down, but I heard it is still very popular. Am I wrong?
86.
by dandaka 7mo ago
Quite a few reasons: 1/ Easy to start, hard to master. 2/ Good balance between being still athletic (you have to run), while not punish less-fit players a lot. 3/ Social component. In Portugal you must get a beer after each m
87.
by dandaka 7mo ago
Love the design and implementation. Even small details like cage on side!
88.
Show HN: Traul – Local-first search engine for all comms for AI agents(github.com)
3 points by dandaka 7mo ago|0 comments
89.
by dandaka 7mo ago
human instructors for wealthy, robotic AI for the rest of us
90.
by dandaka 7mo ago
1/ KYC is pricey, and users might not want to pay for it 2/ Spammer can hire real people to farm accounts I think this idea might work if we - create reputation graph, where valuable contributors vote for others and spread reputat

More ›