endymion-light

6 ms·

31.
by endymion-light 5mo ago
yes - the vast majority of labs offer a whole host of training material for users to understand how LLMs work. There's entire course websites created by each of the major vendors specifically to understand how LLMs work. Here's a
32.
by endymion-light 5mo ago
i mean it's more like "AI can solve this one problem, but it needs X, Y, Z, because it's not a omnipotent god entity" "I tried it without any of those things and it didn't work - this is worthless tech!" I
33.
by endymion-light 5mo ago
Great point - and i'd love a study to address that. If the study pointed out that X services sit perfectly within the analysis found, I think that would be a fantastic study that would be enlightening & useful to show.
34.
by endymion-light 5mo ago
That's true - I suppose i'm just dissapointed that this study hasn't seemed to include those within any analysis. Being able to point out that the top 100 calorie counting apps on the app store return similiar results to simp
35.
by endymion-light 5mo ago
You can commit statistical analysis on frontier models and still use commercial applications as an identifier & comparison. Criticism is not vitriol - it's possible to make a wider point about being taken aback by the lack of educa
36.
by endymion-light 5mo ago
But I don't see them using those commercial services in this study - instead, they're using frontier model companies? Is Gemini advertising that you get a realistic calorie count from a picture? Maybe so - in which case i'd t
37.
by endymion-light 5mo ago
That's a good point - and I think there's a wider core lack of knowledge outside of the bubble. I suppose I just expected this study to be a little less 'water is wet' which made me dissapointed, but that may be coming a
38.
by endymion-light 5mo ago
I suppose i'd much rather a study analyse the apps in the app store that are attempting and claiming to do that kind of thing - rather than the base model they might be using.
39.
by endymion-light 5mo ago
I don't believe the authors of this study are stupid. If there are apps targeting people with diabetes that claims to count your carbs with AI, why haven't those been analysed? That would be a far more effective claim. I based the
40.
by endymion-light 5mo ago
There's an incredibly serious lack of education with how LLMs & carb-counting works. This entire article would be better suited to astrology.com than hackernews. When I opened it up, I assumed the author would have at least attempt
41.
by endymion-light 6mo ago
I've slowly began to write about abandoned side-projects. It's actually incredible how much you end up re-picking back up. A gaussian splat converter that I made and abanonded became incredibly useful a few months later when I nee
42.
by endymion-light 6mo ago
I really dislike the default AI slop css - if you're going to do this - please have a design language and taste ideas beforehand. It can help so much in refining the look. Genuine piece of feedback, as soon as I see those gradients + q
43.
by endymion-light 6mo ago
yes. if instagram started performing intensive JPEG compression that made photos choppy and unpleasant, I would consider that an intentional degredation of the software.
44.
by endymion-light 6mo ago
I like the concept of this, but it does feel horrific to share on most social media. I already share a lot of work I do in local dev chats etc - but there's something about the X algorithm that makes sharing anything feel terrible. Alt
45.
by endymion-light 6mo ago
Thank you for this, i think it's fantastic! I've got some open source AR work i'm doing, it's given me some inspiration to build something using it!
46.
by endymion-light 6mo ago
Honestly, I personally would rather a time-out than the quality of my response noticably downgrading. I think what I found especially distrustful is the responses from employees claiming that no degredation has occured. An honest response o
47.
by endymion-light 6mo ago
I'm not sure how much I trust Anthropic recently. This coming right after a noticeable downgrade just makes me think Opus 4.7 is going to be the same Opus i was experiencing a few months ago rather than actual performance boost. Anthro
48.
by endymion-light 6mo ago
Good points - i'll definitely look into switching - just wanted to reflect on what's causing it to stay so popular.
49.
by endymion-light 6mo ago
it's a joke about the quality of samsung tv's rather than a serious comment - i should have said a perceptron could hack a samsung tv
50.
by endymion-light 6mo ago
While cool and slightly scary news - Samsung TV's have been incredibly hackable for the past decade, wouldn't be surprised if GPT2 with access to a browser could hack a Samsung!
51.
by endymion-light 6mo ago
I'm making more of a pragmatic point. While ethics of companies are important, i'm still using OpenAI, Anthropic, Microsoft, Apple etc, so I definitely accept a trade-off between morality and ease-of-use. Currently i've found
52.
by endymion-light 6mo ago
I'm sorry, on a mac, Ollama just works. It lets me use a model and test it quickly. This is like saying stop using google drive, upload everything to s3 instead! When i'm using Ollama - I honeslty don't care about performance
53.
by endymion-light 6mo ago
Yeah - really struggling to understand why people are not grasping this point. Yes, Easter Island was deforested far earlier - but you wouldn't compare the steam engine's capability in resource extraction compared to what people o
54.
by endymion-light 6mo ago
Disagree for dungeon crawler carl & any of the Bobiverse - while they're fine books, wouldn't class it in the category of interesting ideas, it's just pop fiction. I'd look at the following: Hyperion Book One (For th
55.
by endymion-light 6mo ago
I think there's a bit of a convergent evolution House Of Leaves is similiar to Backrooms in a way - they represent the same kinds of horror but it's more like how Weird Fiction converged and inspired in the same way. There's
56.
by endymion-light 6mo ago
Exponential take-off is great until it stops- genuinely, what are the signals showing any of the large models are performing exponential takeoff and recursive self-improvement? Currently a lot of that appears to be marketing hype to drive u
57.
by endymion-light 6mo ago
Yeah agreed. I think that's partially why I find AI still useful - It acts as a filter against a lot of the listicle style content. I then can curate and look at specific blog writers I follow for actual content. But honestly, if googl
58.
by endymion-light 6mo ago
I view a lot of the AI/Bot internet to be slightly a false misnomer. Even before ChatGPT, the degredation of online content was already happening - SEO farms, worsening google search. Most articles you'd find online would be payw
59.
by endymion-light 7mo ago
This is really cool - i'm going to try it on my old disseration.
60.
by endymion-light 7mo ago
close enough - welcome galactic encyoclopedia

More ›