Great comment. Wish it was the top one so I didn't have to read through the others to get here.
Do you have thoughts as to how the courts would debate the deletion that you could present on a similar intellectual plane?
153 karma · joined July 12, 2026
Do you have thoughts as to how the courts would debate the deletion that you could present on a similar intellectual plane?
The simpler issue is that "mimic" doesn't apply to LLMs in the domain of biology. That would be like saying a digital recording of a bird's flight mimics the bird.
Whether or not the plumage on the bird is red (in both life and the video) is an entirely different question.
Maybe I’m misinterpreting "early on." I guess their point holds for the first 500 words or so.
The details matter. If you're a single person living alone at an income of $70,000.00, you're below the median household income and "picking up" other people's subsidies, but that's not indicative of the country.
[1] https://www.healthcare.gov/lower-costs/
[2] https://www2.census.gov/programs-surveys/demo/tables/p60/286...
From there, fine dresses became something affordable to people who never could have dreamed of them. ChatGPT is simply one of these machines, producing "deep research" as a fine dress; just an example.
Chatbot subscriptions will never be ubiquitous household and industry mainstays. I'm asking bigger.
Or maybe they will! I know that if I had something to yap to, that responded quickly, had perfect (better than Whisper) understanding, and remembered everything perfectly, I'd journal with it. This isn't a reality yet. Models are too slow and expensive for this to work with tools like OpenClaw, Hermes, etc.
There's always room for improvement, though. I suspect a tool will emerge for highly detailed OCR that implements a nested bounding-box-based multi-scale approach, effectively OCRing small sections at a time and then gradually compiling them by expanding the surface area using the bounding boxes.
I've thought a lot about implementing it anyway.
edit: I see you're asking about the block labels. Leaving the comment in case someone finds it interesting.
I really wouldn't know, though. Anthropic models barf out copyright issues for my use case, so I'm unable even to benchmark them. It's a common problem when you're scanning public domain books. Mine are reference texts often cited.
As another user pointed out, it's surprisingly random (task-specific). Llama Scout outperformed Gemini Flash 2.5 on a benchmark I built at the time. I didn't include an OCR models.
Mistral might indeed be the best OCR-specific model for my task, now that you ask. Funny. It's so bad at my work that I didn't register it might be the best in its category. This is just based on vibes from my single scan.
Nothing special about this model for overly-detailed work like mine.
It's been a while since I last tested (and discontinued my subscription), but the "pro" models from OpenAI dominate. Not surprising, given the price difference, but it would be nice if an OCR-specific model could perform better. It's worth mentioning that even the highest-end models do a pretty poor job with intricate text like mine.
You pretty much need $30/m subscriptions for historical data. Not that it couldn't be worked into this.
The best feeling has to be in the CLI, in my oppinion. A low-stakes project where you can let ultracode runaway from you in the CLI is the most fun.
Tabs for long-running often-compacted sessions is a difference I've noticed in emergent UX.
I don't know how to articulate it, that's just my best shot. I'm baffled by my poor decision-making and assume the author is guilty of the same faulty reasoning.
I think people fall into the trap of thinking things are too good to be true (so they must not be true). This results in abandoned relationships that would have been better than 99% of those from pre-online-dating times.
This has been true for me, anyway. Every time I've entered the online dating scene, I've invariably realized 3 months later that one or two of the women I turned down for second dates were great matches.
I'm glad I can't, but I'd like to be informed. Is it prompt engineering or just some model/harness combination?
Edit: I misread the chart! It would indeed be F#
I've got some education in materials engineering; it'd be trivial to drop a curve (line) for optimizing the language selection, and Python obviously comes out on top.
The author doesn't even mention it. That discredits the whole article as far as I'm concerned.
There's a large market, very large, who want the best regardless of what it costs. Probably a large enough market to keep that domain of research afloat (as opposed to shifting research manpower to cost cutting).
The reasoning is just that the marginal cost of AI is very secondary to fixed costs of the businesses themselves; it's not an excuse to sacrifice performance.
I've been able to accomplish incredible feats (for myself) since GPT-4, so model intelligence is secondary.
I think I disagree with your basic point, though what you say about normal vs. hard is funny. I'm fortunate enough to work with a UI/UX pro, and these things have right answers; they're incredibly obvious after they've been explained to you.
Normal varies so much.
As a "cozy" gamer, I want the standard setting to be something I can play without having to switch my context to to the internet for help or a guide. This was one of many triumps of Baldur's Gate 3: the difficulty settings make sense in how they translate to gameplay, and the default is fun.
To your point: I think you're simply mistaken that they offer a good scraper-detection solution, but I'm all ears.
For those inclined to multivariable calculus, do your dopamine crow brain a favor and set pencil to paper.
It's super-intuitive because everything's a circle. The equation for calculating volume is beautiful, and figuring it out is fun.
Just thinking about a wave equation for the graphic is interesting
Scraper "attacks" don't take down our robot-specific server very often; it's safe for us to take heavy-handed approaches that sometimes redirect users there. 99% (made-up high number) of the time, the misdirected users don't realize anything is amiss.
Start by analyzing your traffic, specifically user agents. Look for "robot" or even "bot" in the user agent and load balance those to a robot-specific server. This can all be done within Cloudflare. The only code is the user agent condition. Note: I'm very open to input here if anyone reading notices that we're shooting ourselves in the feet. Based on our analysis, the remaining traffic is a good picture of our human users.
We have loads of other conditions, mostly balancing specific IP ranges for entities when we know exactly who they are, but this is a good start.
We direct scraper traffic to a bot-specific server using Cloudflare's load balancer, slowly analyzing traffic and adding conditions one at a time. No accidental scraper DDoS in a long time.
Most scrapers are relatively honest in some way shape or form.
I was curious: professional/scientific/technical saw little to no growth; the areas that would be most affected by AI.
I've noticed or experienced absolutely none of this. The closest thing I can relate to is that I operate on a different intellectual plane from some of my blue-collar friends. This causes a divide and probably limits intimacy, but does not result in negative relationships.
Regarding the dating world, it impresses people that I'm an SWE. My role is a boon, maybe because it signals intelligence; I'm not sure.
Strongly disagree here. Would encourage you to try to reframe some of that. Sorry if that comes across as patronizing.
"Bullshit Jobs"