Viewing profile — ktzar
ktzar
HN member- Joined
- Tue, Feb 12, 2013, 9:16 PM UTC
- HN karma
- 519
- Public activity
- 164 items
- HN profile
- View on Hacker News ↗
About ktzar
No profile information was provided.
Recent public activity
-
comment
Comment #49216197
Emily Wilson's translation is way better, more modern and adequate to present-day readers and, probably more context aware.
-
comment
Comment #48952399
Is it me or this is one of the worst and knowingly less informed articles that has hit HN in a while?
-
comment
Comment #48940299
Hallucinates on the first question I ask, as 90% of these models that try to take shortcuts.
-
comment
Comment #48618171
it's one of my favourite interviews questions and very few people get it right.
-
comment
Comment #48553043
For someone so mindful of efficient software and use of energy, it's showing that images on that post are 5MB PNGs...
-
comment
Comment #48525642
Such an old problem solved very elegantly. Congrats. Remember the days of MySQL Workbench and how clunky it was.
-
comment
Comment #48492432
A good way to think about it is finding how much it'd cost to buy and run a GPU that runs a model at around 100tk/s ("thinking" agents are not viable otherwise). The figure mention…
-
comment
Comment #48409937
compatible with Cardputer?
-
comment
Comment #48335359
Pandoc is such an amazing piece of software. I used it to format my novel and made it part of a GitHub action to produce all the formats I required. I wasn't aware of templates, bu…
-
comment
Comment #48093131
I don't think we're in the early ages... LLMs technology has essentially stagnated since GPT3.5, we just have bigger models that can handle more context. We're trying to cope for t…
-
comment
Comment #47779711
Writing your own storage is a great way to understand how databases work (if you do it efficiently, keeping indexes, correct data structures, etc.) and to come to the conclusion th…
-
comment
Comment #47114935
I LOVE this, but I am still love the Sunday night family bonding moment of planning the week in the pen and paper weekly planner that also costs us $10 a year.
-
comment
Comment #46853479
because we have an alternative that we humans can fix. The problem with AI is that it creates without leaving a trace of understanding.
-
comment
Comment #46841834
Didn't know about Grokipedia, I've just opened an article in it about Spain, scrolled to a random paragraph, and the information in it is plain wrong: From https://grokipedia.com/p…
-
comment
Comment #46748049
I didn't know about Tansu and probably would not use it for anything too serious (yet!). Bus as a firm believer of event sourcing and change of paradigm that Kafka brings this is c…
-
comment
Comment #46510837
Just that all of those activities you mention feel like a useless life compared to spending time with your own children in a house big enough for everyone to have their space, but …
-
comment
Comment #46474716
There's no more proof that any Venezuelan election's results has been tampered with than with any US election. The state of Venezuela's state is sad, and so is the fact that millio…
-
comment
Comment #46468471
I really enjoy using text.npr.org from my Kindle / Kindle Scribe. I'm really thinking about setting up a self-hosted RSS aggregator site that's Kindle-friendly.
- story
-
comment
Comment #46347283
I developed a big chunk of my Scumm games decompiler in London's central line. I was lucky (or unlucky) enough to go far enough each day to always hand an empty seat and enjoy 30 m…
-
comment
Comment #46343171
It's the same hunch we all have when we think we're going to learn something by watching tutorials. We learn by struggling.
-
comment
Comment #46140510
I think you hit the nail in the head there. There's absolutely nothing we can do with AI that we can't do without it. And the level of understanding of a large codebase that a soli…
-
comment
Comment #45874839
Consequences of having early access to ChatGPT and getting AI knowledge debt?
-
comment
Comment #45357848
And most importantly, we're paying with our brain and skills degradation. Once all these services stop being subsidised there will be a massive amount of programmers who no longer …
-
comment
Comment #44783665
Also, the way models are evolving (thinking process, llms waiting for interactions with external entities via MCP, mixture of experts, ...) are making "useful chatbot responses" wa…