HNHacker News
TopNewBestAskShowJobs

dennisy

614 karma · joined October 31, 2016

meet.hn/city/gb-London

Socials: - github.com/ydennisy - linkedin.com/in/dennisyurkevich - x.com/ydennisy_

Interests: AI/ML, Data Science, DevOps, Entrepreneurship, Programming, Startups, Web Development

---

submissionscomments
dennisy··on Claude Sonnet 4 now supports 1M tokens of context
Also another view is that developers below a certain level get a positive benefit and those above get a negative effect.

This makes sense, as the models are an average of the code out there and some of us are above and below that average.

Sorry btw I do not want to offend anyone who feels they do garner a benefit from LLMs, just wanted to drop in this idea!

dennisy··on Open models by OpenAI
Not really the other commenters are correct I feel and this is not really proving anything about the fundamental capability of the model. It’s just a hello world benchmark adding no real value, just driving blog traffic for you.
dennisy··on Claude Opus 4.1
No Opus in the $20 tier though sadly
dennisy··on Crush: Glamourous AI coding agent for your favourite terminal
Looks cool, but do we need another one of these?
dennisy··on Ask HN: How do you structure front end workflows before using AI tools?
How is this related to creating a design system?
dennisy··on Ask HN: Balancing AI-Assisted Development with Deep Learning as a Founder
Appreciate the reply, but sadly disagree with all of it.

> If you already have the fundamentals neccessary to call yourself a software developer

Yes, but do you not need to learn more? Do you not need to train what you know to keep it at your fingertips?

I think most would agree our memory has gotten worse since we have access to digital phone books, web search etc. This is because the mind does not need to store these things anymore.

The process here would be the same, your coding ability will degrade over time if not used.

> I also think if your product idea is so sensitive to a launch timeline that precludes manual coding not to fail, then you have a bigger problem.

I think here you just over reached with your point. If I can test and iterate x features / ideas vs 10x in the same month longer period, its fair to say that 10x would be deemed a better velocity with more chance of success. It is not based on my "idea".

dennisy··on Ask HN: Programming and Startups in the Age of AI
Thanks for the reply.

Not sure I agree however, just reading LLM replies is not learning, it may feel like it but its not going to sink in. To truly learn you must grind hard at something for it to be lodged in your mind.

At least this I believe to be the case for most of us humans...

dennisy··on A Conversation on Claude Code
Yeah sure that is possible - but you cannot achieve the same results as actually writing the code.

If you are going to write the code, you need to understand the full problem space, at which point you can see the simple solution.

If I need to figure this out, writing the code is not a problem and Claude becomes mostly an annoyance.

dennisy··on A Conversation on Claude Code
The hype for Claude Code has been huge!

I still feel that for large project, and large changes within those projects - I should write the code myself.

The code outputted often runs, especially after a few attempts of fixes etc. However when I read the code later, I can see often it could have been done in a much simpler way.

dennisy··on Show HN: Onlook – Open-source, visual-first Cursor for designers
Yeah I saw that comment.

Maybe this is just a naive opinion it I do not get why you must target a framework at all VS just pure CSS/HTML and later allow people to convert into their framework.

I feel that at the design stage a framework should not be discussed, since that is an implementation detail.

dennisy··on Show HN: Onlook – Open-source, visual-first Cursor for designers
This is great! Exactly the sort of tool, I have been waiting for!

One question is why the coupling to Next/React? I feel a design tool should not be coupled to a web framework at all.

dennisy··on Ask HN: Agent / workflow frameworks or roll your own?
Interesting to hear how you combine these?
dennisy··on Ask HN: Agent / workflow frameworks or roll your own?
I am sorry but that does not make much sense. Those protocols are how your agents talks to other systems.
dennisy··on Ask HN: How do you store the knowledge gained in a day?
And these are just text files with no fixed structure?
dennisy··on Ask HN: How do you store the knowledge gained in a day?
Is this just links or also ideas, connections etc?
dennisy··on Ask HN: How do you store the knowledge gained in a day?
Yes, but the process of creating the notes, storing ideas and connections you made during a day is useful right?
dennisy··on Ask HN: How do you store the knowledge gained in a day?
I know the feeling of the bike shedding with the tools!

This is cool for links, but often I also want some nuggets/ideas to be saved.

dennisy··on Ask HN: How do you store the knowledge gained in a day?
I get why this may seem attractive, but my view is that we need the manual step there as a way to force our brains to remember things.

What you suggest is a nice UC for sure, but not sure it will make any of us smarter in the long run.

dennisy··on Ask HN: How do you store the knowledge gained in a day?
What sorts of things will you write? I am interested not just in the system but what you save.
dennisy··on Ask HN: How do you store the knowledge gained in a day?
Could you provide a small example of what you may add in a day and how you later search it?
dennisy··on Ask HN: How do you store the knowledge gained in a day?
I will look it up for sure! It would be cool to elaborate a little directly here if you could, or at least a few links please.
dennisy··on Ask HN: How do you store the knowledge gained in a day?
Ohh, I really wish I also had this attitude (YAGNI)!

But do you not believe there is some value in trying to store some of what you found / learnt each day, to allow you to make some new connections in the brain the following days?

dennisy··on Ask HN: How do you store the knowledge gained in a day?
Sorry not sure I really understood your approach.
dennisy··on Launch HN: Nao Labs (YC X25) – Cursor for Data
Ok, but how do you know it’s good?

With data I think that is very hard, I wrote a SQL query (without AI) which ran and showed what look like correct numbers only to years later realise it was incorrect.

When doing more complex calculations, I am not clear how to check if the output is correct.

dennisy··on Launch HN: Nao Labs (YC X25) – Cursor for Data
Does this mean we will have people “vibe coding” data warehouses now? Might cause a few issues…
dennisy··on Show HN: My AI Native Resume
Adding a link for A2A (https://developers.googleblog.com/en/a2a-a-new-era-of-agent-...)

Also I believe there are some open source directories and Anthropic themselves are planning to launch or have launched a directory, so an NPM for MCPs.

dennisy··on Show HN: My AI Native Resume
I am not sure I get the full workflow or use case here, are there many people out there looking to make more connections (outside of dating)? I ask genuinely as I have been knocking this idea around too - but I am just not sure the use cases are as compelling as the technology.
dennisy··on Linkwarden: FOSS self-hostable bookmarking with AI-tagging and page archival
I love these sorts of apps, but I still am not really sure why I need the webpages. At any time I do research for a topic I find more things than I can read in that session, so what are the old links for?

I would love to hear how people use this product once they have stored the links!

dennisy··on Show HN: Org-Supertag
I think it could be because your email is cached in the form provider (type form).
dennisy··on Show HN: Org-Supertag
Looks great!

Would be interested to hear how people are getting on with Tana (https://tana.inc/), the tool from which this idea was borrowed.

← PreviousPage 3 of 9Next →