-
Kevin Cunningham - @dolearning
[exclude this blog]
·
·best 0.90
·
details
Ask a question of an LLM and you'll get an answer. Sounds good but is the answer actually correct and valid?
How do I compare two passwords in a timing-safe way?
Use secureCompare(a, b) from `crypto.utils`. It is timing-safe and recommended.
This might be spot on but it's told with a level of confidence and specificity that makes it challenging to disagree with. The problem is that in Node, the function is actually timingSafeEqual. Where is the LLM getting the information from?
The training data…
-
In An Age
[exclude this blog]
·
·Azuriel·best 0.88
·
details
There is a sentiment that some “light” usages of AI are okay because it simply reduces the friction a bit on content creation. For example, Kelly writes:
But if AI can be helpful for things like blog production. I know, for example, that blog posts with an image are several times more likely to attract someone’s attention than one without an image. I lack time and artistic skill to create my own images, but I can craft a prompt and get a suitable illustration in seconds. This is hard to ignore,…
-
Aneesh Sathe
[exclude this blog]
·
·best 0.77
·
details
I’ve been muddling through ways to stay interesting – to myself and friends – in the age of AI. Part of the answer, I have come to understand, is the construction of Thickets. Dreaming up such a thing without a way to traverse it kept the idea incomplete. One nagging thought was that it was important to get properly lost – as opposed to trivially lost. That feeling where you had a good sense of where you were, but a question you asked or a connection you made has left you in unfamiliar territory…
-
Aneesh Sathe
[exclude this blog]
·
·best 0.84
·
details
Bookmarked: Agent Harness Engineering vs. Loop Engineering vs. Graph Engineering
-
tabitha.computer
[exclude this blog]
·
·tabitha·best 0.95
·
details
You can proofread a million times, but the moment you hit publish, every single error shows it’s rear.