uludag

14 ms·

151.
by uludag 2y ago
I don't think people finding LLMs useless is a good representation of the general sentiment though. I feel that more than anything, people are annoyed at LLM slop. Someone uses an LLM too much to write code, they create "slop,&q
152.
by uludag 2y ago
But according to the author, apparently bringing this up isn't helpful criticism. I'm curious what peoples thoughts are of what the future of LLMs would be like if we severely overshoot our carbon goals. How bad would thinks have
153.
by uludag 2y ago
I still think this could worthwhile though for these reasons. - One "quality" poisoned document may be able to do more damage - Many crawlers will be getting this poison, so this multiplies the effect by a lot - The cost of genera
154.
by uludag 2y ago
leetoced/AoC-like problems are probably the easiest class of software related tasks LLMs can do. Using the correct library, the correct way, at the correct version, especially if the library has some level of churn and isn't extre
155.
by uludag 2y ago
My thoughts exactly. I honestly think that the majority of SWE work is completely unoriginal. But then there are those idiosyncrasies that every company has, a finicky build system, strange hacks to get certain things to work, extensive we
156.
by uludag 2y ago
Thanks for sharing your thoughts! I too like to peruse the things that neovimmers are up to as they come up with some pretty neat ideas. There does feel like there is a productive synergy between Emacs and neo/vim.
157.
by uludag 2y ago
I'm always curious how poisoning attacks could work. Like, suppose that you were able to get enough human users to produce poisoned content. This poisoned content would be human written and not just garbage, and would contain flawed r
158.
by uludag 2y ago
> The better terms won out... I actually don't think they won out. Mindshare maybe, but I use kill and yank on macOS all the time. In fact, since they are enabled by default, I use Emacs keybindings all the time and everywhere on my
159.
by uludag 2y ago
The percentage of the pie is smaller, but the pie has grown so much larger that in absolute numbers, there may not be any decrease. > And also potentially shortsighted. Looking at the recent releases, it feels like the vscode enshittific
160.
by uludag 2y ago
> EmacsConf feels like a nice, cozy get-together where people share the cool things they've been working on and thinking about. Emacsconf was indeed executed great this year, and nice and cozy were my feelings too. It's interes
161.
by uludag 2y ago
Good point, often times it is the miraculous that attracts the most detractors and a benign or useless technology would be ignored. I guess I meant something more like a widespread negative reaction means something's probably not right
162.
by uludag 2y ago
I think I've started to prefer text with spelling and grammatical errors now, to a certain degree at least.
163.
by uludag 2y ago
I think saying that the author is stuck in his ways and is just fighting change is a bit too simplistic. - All major tech companies becoming "AI first" - The websites we use are still slow and buggy - "AI" feature bloat
164.
by uludag 2y ago
But in all honesty, if LLMs were really that miraculous, we wouldn't have such detractors. Like, I don't remember any widespread anti-search-engine sentiment when Google came out. I can name many technologies that don't get s
165.
by uludag 2y ago
It's actually even worse than that: the current trend of AI is transformer-based deep learning models that use self-attention mechanisms to generate token probabilities, predicting sequences based on training data. If only it was somet
166.
by uludag 2y ago
I feel like this is a case of the no silver bullet principle. Like, going from a legacy/pre-modern/old language to modern, dynamic to static typed, etc. isn't going to magically make your projects succeed. Maybe these failed
167.
by uludag 2y ago
Oddly enough, for 2025 I would say writing. I would normally think that with the rise of LLMs, writing would be come obsolete. Interestingly though, I think LLMs make good writing more valuable and at the same time more rare.
168.
by uludag 2y ago
Work at a "tech shop," even old-school ones, most likely involves extremely varied types of work, much more than name-refactoring, for which something like IntelliJ indeed would be the best tool. Just because one use-case of one
169.
by uludag 2y ago
I'm pretty sure the plan has never been to just make these tools that make us more efficient. If AI stays at the level it's at, it would be a profound failure for companies like OpenAI. We're all benefiting from the capital b
170.
by uludag 2y ago
I think Emacs definitely has a certain uncool vibe about it. Like, this comedy video from "Programmers are also human" sums it up pretty well: https://www.youtube.com/watch?v=urcL86UpqZc . And I think this mental i
171.
by uludag 2y ago
I have similar feelings though maybe a bit more optimistic take. Obviously the AI hype train hasn't taken us to anywhere objectively better. Software has in no way become less buggy (if anything it feels worse in the past few years) a
172.
by uludag 2y ago
I feel like this free plan is just enough to get someone hooked and require them to upgrade to paid. Like, imagine GPS navigation wasn't widespread and there was a paid service that gave you 20 free trips. Eventually your normal navig
173.
by uludag 2y ago
ok, I think I understand what you meant better. I guess my mental model of UI has some sort of utility baked in but I see how this isn't necessarily the case. I would agree with this take actually. It is deffinitely is a jarring expe
174.
by uludag 2y ago
Sounds like this could be the premise for a chapter in the book "Invisible Cities": I shall now explain to the city of Fransisca, the blue city upon who's golden shores the azure waves drift over, and above which a sapphire s
175.
by uludag 2y ago
From what I've noticed, despite the label "outdated", I can't reliably notice any difference in utility between these categories. neo/vim users (which I certainly don't use) are at least as effective using thes
176.
by uludag 2y ago
Firstly, as many commenters have mentioned, I don't see AI taking jobs en masse. They simply aren't accurate enough and they tend to generate more code faster which ends up needing more maintenance. Advice #1: do work on your own
177.
by uludag 2y ago
Reading this, I realized a reason why I try to move as much of my computing life into Emacs as possible: because having full control over a, perhaps inferior, Elisp application does result in a better experience, as I have full control over
178.
by uludag 2y ago
I was reflecting on old technologies that were supposedly revolutionary at the time, like palm pilots, and what value they really provided. Like, I can imagine a someone at that time using a palm pilot and claiming that it's making th
179.
by uludag 2y ago
I don't find it controversial. I think it's a decent heuristic that's simple to implement. It would be pretty trivial to extend this with other heuristics, like how often commands other than self-insert-command are ran, their
180.
by uludag 2y ago
Interesting to put this besides Hofstadter's law, which looks to be also real, which states "It always takes longer than you expect, even when you take into account Hofstadter's Law." I don't think these need to be

More ›