MisterKent

5 ms·

1.
by MisterKent 7d ago
There's less gray area with driving wss the point. Legal or illegal, efficient or inefficient. And whatever gray area remains is generally "good enough" compared to human drivers on the road.
2.
by MisterKent 14d ago
Good. The more I use AI the more I feel like it will drag humanity into a local maxima that we'll never be able to escape from.
3.
by MisterKent 26d ago
Based on the comments here and the content of the post... I sort of agree. But maybe less so directly with their gatekeeping. But maybe we should formalize closer to "real professions". And if we think college degrees don't m
4.
by MisterKent 1mo ago
Hey Ryan, just want to congratulate you on this. I've _loved_ solid since well before it's v1 release, and use it in all my side projects. If anyone hasn't tried it I highly recommend it. Simply put: Solid makes the "dum
5.
by MisterKent 1mo ago
"Everyone's doing it" presented with no evidence is just you trying to make yourself feel better for: - using from - driving a Tesla - voting for Trump It's a really, really easy line to draw in the sand: don't supp
6.
by MisterKent 2mo ago
Maybe just allow dragging independent of validating / submitting. So I can build the full page out and then validate?
7.
by MisterKent 2mo ago
Not a bad concept, but there's an unwritten rule in most puzzles that the clue and answer can't overlap. Really messed with me for shark tank. Meanwhile, and I say this as someone who plays a lot of puzzles, this feels distinctly
8.
by MisterKent 2mo ago
I've said this before on my blog, but the new metaphor I've been using is drug development. It costs billions to develop a drug (model) and pennies to produce the actual pill (tokens). The government regulates the drug industry to
9.
by MisterKent 2mo ago
Right. If you are used to docker compose and don't care about redundancy or authoring your own stuff. Just setup storage and then use helm install. It feels like an app store at that level. Grow from there as needs arise.
10.
by MisterKent 2mo ago
I learned it for this setup. I do zero devops at work, there's a separate team that handles that (and even then, they're not on kubernetes). I learned it because it seemed like the right way to do things and fun. Cut myself a few
11.
by MisterKent 2mo ago
Lot of kubernetes hate here, which is surprising. I run a little 3 node cluster and besides the hardware issues I had (long story), it has been rock solid and dead easy to setup. Talos + longhorn + fluxcd (optional), is super nice. And ever
12.
by MisterKent 3mo ago
Is this really where we've landed? I refuse to believe that any of this markdown insanity will continue indefinitely.
13.
by MisterKent 3mo ago
That's... A lot of words to not say much. I actually read this, and still have basically no idea what your point was. But you sounded very pretentious and AI-ified while making it.
14.
by MisterKent 3mo ago
Do we really need to keep pandering to the people who block JavaScript and fonts? At some point it's on them to live with their choices. And just from my small sample size (I doubt there's a large sample anywhere), those people ar
15.
by MisterKent 3mo ago
I ran Ms-01s with 100GBE, copper DACs in my kubernetes cluster. Killed the NVME drives in that tiny box. I'd bet the same issue doing this with FW. And I wasn't even pushing 100GBE very hard at all, it was mostly for fun. AI + 100
16.
by MisterKent 3mo ago
That was my experience with Fable as well. Pulled my extremely complex project that I could squint and see was possible, but actually put mathematical concreteness to things in a way I could only intuit. On the flip side, visualizers have a
17.
by MisterKent 3mo ago
Apple is actually interesting. They are one of the few companies with a chip / PC play with real power AND basically no play I'm the hyperscalar market. That means they're actually incentivized at least short term, to benefit
18.
by MisterKent 3mo ago
If Fable is only for US citizens, presumably that will apply to enterprises as well. I used fable on some difficult stuff and it was surprisingly good. It's safe to say that models aren't going to get worse. Does that mean: US cit
19.
by MisterKent 3mo ago
Sorry. This is addressed at the 99.9%. the them is the 0.1% of people that are responsible for the 99% of the bad news the article mentioned. Didn't think that was needed to be pointed out.
20.
by MisterKent 3mo ago
Things are worse than ever. And these types of opinions are just Ostrich strategies. This worked fine 20 years ago, and would be a wildly inappropriate and irresponsible way to live today. Yes, a lot of the news is sensationalized and blown
21.
by MisterKent 3mo ago
They want people to believe they have massive models, that is effectively their moat at this point. Because if they don't imply that size is needed for every task, they'll end up tanking their valuations. https://blog.n
22.
Closed-models are do not (financially) scale(blog.nilesh.io)
1 points by MisterKent 3mo ago|0 comments
23.
by MisterKent 3mo ago
If this makes Elon a trillionaire then he's not the first. All those NFT bros listing their stickers for trillions each have him beat by orders of magnitudes.
24.
by MisterKent 3mo ago
I keep seeing this sentiment in general and in this thread. I wrote about it here too: https://blog.nilesh.io/post/llms-and-jobs Just because people have a vague idea of what someone's job is and a deep understand
25.
by MisterKent 3mo ago
Now I can produce slop without AI.
26.
by MisterKent 4mo ago
Of course, but that's kinda the point I'm making here: It's a tool not a replacement. Job losses due to productivity may be inevitable, but that's not the same as being replaced by AI, that's being displaced.
27.
The LLM Job Paradox(blog.nilesh.io)
3 points by MisterKent 4mo ago|4 comments
28.
by MisterKent 4mo ago
Wow. Didn't realize OAI was astroturfing hacker news now...
29.
by MisterKent 4mo ago
Software engineering is the practice of taking a concrete problem, converting it into a set of rules (a world model), and then converting those rules into something a computer can execute for a human user. As long as AI cannot dynamically g
30.
by MisterKent 4mo ago
I imagine any real experts in art would just recognize the painting and so didn't make the "cut"

More ›