nadam

15 ms·

1.
by nadam 2mo ago
Does anyone know why these Claudisms exist? Are these kind of expressions a result of some kind of 'evolution', so they are more effective in chain-of-thought reasoning that other expressions? Or the preference of people doing som
2.
by nadam 8mo ago
You put the "agent" word into apostrophes as if I use it as a marketing buzzword. No. An agent is an LLM in a loop with memory usage, with file system access, etc, which is usually more effective than just using an LLM as is, espe
3.
by nadam 8mo ago
Maybe. But I am working with these agents and I see the sophisticated patterns using LLMs, using ground truth data and human experts working together. It could be much more effective than 'patient asking chatgpt' vs. 'general
4.
by nadam 8mo ago
I am amazed at the false dichotomy (ChatGPT vs. Doctors) discussed in the comments. Let's look at it from the perspective of how a complex software is developed with a team of software engineers and with AI assistance. There is GitHub,
5.
by nadam 11mo ago
I find it amusing how people take this 'leveling game' at big companies so seriously. The title they get at these companies become their identity. They live by the rules the company imposes on them. Amusingly almost the opposite o
6.
by nadam 2y ago
I was just thinking about what AI assisted coding brings to the design discussion, especially as AI become more and more powerful and we rely on it more and more. You still want to make things modular and easy to understand so that AI under
7.
by nadam 2y ago
What I am starting to internalize nowadays is that working faster is useful, but what can really change one's life is learning faster.
8.
by nadam 2y ago
Location: Budapest, Hungary Remote: Yes Willing to relocate: No Technologies: I have the most experience in C++. But there are several others, like, Java, Typescript, Angular, etc... Résumé/CV: https://docs.google.com/d
9.
by nadam 2y ago
Hmm, I think 'fail fast' and 'embrace the grind' are popular and somewhat contradictory advices. Which is better? I think 'fail fast' is (or was?) a bit overhyped so I tend to err on the side of 'embrace t
10.
by nadam 2y ago
"But when you have to scale the work, then the code has to be easy to read, easy to extend, easy to test, have extensive test coverage, have proper dependency injection / be easy to mock the 3rd party dependencies, be easy to conf
11.
by nadam 2y ago
"When every supermarket aisle looks like a sea of sameness, when every category abides by the same conventions, when every industry has converged on its own singular style, bold brands and courageous companies have the chance to chart
12.
by nadam 2y ago
An example is the material editor. When dragging a value of a property of a node (like color) you see it immediately in the editor. You apply a material on an object, (regular object, water, Landscape), and while changing stuff in the node
13.
by nadam 2y ago
Re-watching Bret Victor's "Inventing on Principle" was an interesting experience today. At the time of recording (2012) his talk was magical and ahead of its time. Today his principle (immediate feedback to the creator) is fu
14.
by nadam 2y ago
"compare this with enterprise software, which is orders of magnitude more complex than video games in terms of business logic" Maybe this was true 20 years ago, but I do not think this is true today. Game code of some games is alm
15.
by nadam 2y ago
His system prompt is too specific. It worked for me with GPT 4o using this prompt: ---- System prompt: Please note that words coming to you are tokenized, so when you get a task that has anything to do with the exact letters in words, you s
16.
by nadam 2y ago
Or alternatively: 1. Programming in very concrete/practical terms because you do not know how to think in precise and abstract terms (do not know math) 2. Thinking more precisely and abstractly (more mathematical way) 3. Only do some k
17.
by nadam 2y ago
Amazing work, prompt engineering at its finest. One future direction for Arc AGI could be to use not Python, but a much more concise programming language that is more suited for brute-force methods like genetic mutations. The problem would
18.
by nadam 2y ago
I love this, this is super interesting, but my intuition based on looking at a dozen examples is that the problem is hard, but easy enough that if this problem becomes popular, near-human level results will appear in a year or less, and AGI
19.
by nadam 2y ago
> 'The breaking tech news this year has been the pervasive spread of "AI" (or rather, statistical modeling based on hidden layer neural networks) into everything' Calling current generative AI tech as 'statistica
20.
by nadam 2y ago
I think it is important to discuss the notion of accidental complexity and essential complexity here. If your organization's strength is that you have world class engineering essential complexity is your friend: a problem domain with b
21.
by nadam 2y ago
It is too early to tell for my current project. I know that it is very hard to create a successful game. Certain genres are more crowded than others though. Creating a successful 2D platformer is almost impossible nowadays for example. I am
22.
by nadam 2y ago
It is not just about the financials. It can be a very lonely activity to create something that other people do not care about, or just very very few people care about. (As it is described in the article.) Besides financial reward, social re
23.
by nadam 2y ago
You are right: it could still work as something I enjoy. It is just that I am in a state of my life where I do not want to put too much time into such an activity: I have a greater desire to create things that people want. But I still watch
24.
by nadam 2y ago
The problem with nerd sniping is that it is instant gratification in a way. Always focusing on tech and math problems because you like solving them can lead to neglecting other areas. But maybe learning more about those other areas could le
25.
by nadam 2y ago
Not very theoretical. But with 'nerd sniping' focus I tended to create my own tech unnecessarily, and wasting too much effort to optimize things that do not necessarily need to be that much optimized. Nowadays I try to reuse other
26.
by nadam 2y ago
I try to find joy in creating stuff that some other people find useful or enjoy. Creating something just so that I enjoy the process but society do not value at all is something I have a very bad feeling about nowadays. Like I am wasting my
27.
by nadam 2y ago
Joy with creating something other people enjoy or find useful. For example creating a videogame. But formerly I enjoyed creating the underlying technology (engine programming, which is a nerd sniping too), nowadays I am more interested in t
28.
by nadam 2y ago
"If you want to go into mathematics, doing the mathematics itself has to be the thing that’s the reward" True. In other words mathematics is the biggest 'Nerd Sniping' source ever existed on Earth. Math was my hobby when
29.
by nadam 2y ago
In my perspective it is more believable to get into AI as a young guy without higher education than someone like me: a middle-aged person with higher education with a family and a reasonable corporate software engineering job. For me to get
30.
by nadam 2y ago
The main problem is that as a software developer most of us worked at software product companies all our life and I noticed that often we know more about software product design and product strategy than some of the 'idea people'.

More ›