HN user

Traubenfuchs

3,622 karma
Posts3
Comments2,175
View on HN

Mark Zuckerberg‘s and Meta’s incompetence should have been recognized after the metaverse debacle.

They were lucky with the ad empire he built and that‘s it.

Claude Sonnet 5 22 days ago

I think the cool kids call this "staying away from the vector space of highly skilled security engineers".

Claude Sonnet 5 22 days ago

Completely different results as LLM is non-deterministic.

You'd need to produce this like 20 times by each model and then do 2x20x20 cross comparisons by both models and ultimately distill the 2x20x20 comparison results into two reports of how they differ.

In this non deterministic computing future, everything else is voodoo, feelings and "vibes".

Claude Sonnet 5 22 days ago

This slot machine has access to your bank account and can decide how much to play on its own!

Claude Sonnet 5 22 days ago

It's really bad when you let opus do investigations on broken java or infrastructure stuff. It starts decompiling .jar, sometimes multiple versions of the same dependency, reading every single kubernetes/terraform file and loading all the logs and info kubectl offers.

I am still quite confused on the scientific consensus:

Should we double down on renewable energy and solve its issues with lots of batteries or should we invest in next generation nuclear energy?

Both at the same time?

Does anyone know?

specialized hosting

What do you mean?

My private next.js fullstack slop runs dockerized on my kubernetes cluster and for auth I use auth0, because I am too lazy to run keycloak or whatever dockerized auth slop is currently en vogue.

It would actually be good for us…

I never understood why software engineers were so excited about open source and teaching everyone to code.

Why aren‘t we more like doctors or lawyers?

There is ALWAYS a need to scale up, upgrade versions, migrate from one database or scheme to another, cut services into more microservices, melt them back into a monolith, integrate more stuff, deprecate other stuff, fix bugs, increase performance, tune caches…

Some things happen for reasons that make sense, others are done for the sake of it. The PO‘s of front end features need to evolve them. What do you do with a PO who says this is done? They don‘t have visibility, don’t get promoted and they don‘t get a bonus. We have created a dashboard that shows that A/B tested users like the new design more by 7%! The next redesign is already planned!

Is that why there are billions dollars wasted in useless Microsoft subscriptions and services?

Microsoft is still simply one of the very best at enterprise dealmaking.

Anthropic is getting here too.

I almost exclusively use claude for all my professional and private needs. In my experience it's really good at adhering to my wishes in regards to sycophancy and pushing back. If you really want to you can tell it to systematically push back on anything where pushback makes sense until it continues with the flow of conversation.

In my first therapy session, the answers were too long and contained multiple questions, spawning multiple threads of conversation. I told it to tone it down and only ever ask one question back, maybe two, if they are related. The answers got too short. I told it to make them "slightly longer" again and reached a sweet spot.

The conversation is yours to form! You need to find the "system prompts" and guidelines to give it that work for you.