suttontom

5 ms·

31.
by suttontom 3mo ago
Your instinct is correct, and in a lot of cases it's true. However, I've heard from enough doctors by now (a cardiologist, psychiatrist, and epidemiologist/former physician) that they use medical LLMs and find them extremely
32.
by suttontom 3mo ago
How did it help?
33.
by suttontom 3mo ago
It's a blog. He's using hyperbole. It's not a Supreme Court opinion.
34.
by suttontom 3mo ago
I'm asking genuinely, is there a connection between housing, education, and healthcare becoming so much more expensive and them also being the three parts of the economy that have the most government interference (in the US)? If so is
35.
by suttontom 3mo ago
I don't want to be cynical, but maybe spending hours every day using Claude has made some of us particularly attuned to picking this up. For some reason as soon as I read "The trap was in app/test/index.js," I insta
36.
by suttontom 4mo ago
You're wrong in lots of ways. Some model cards do show regressions on benchmarks for newer models on specific tasks: https://storage.googleapis.com/deepmind-media/Model-Cards/Ge... This wasn't a new mode
37.
by suttontom 4mo ago
I wouldn't agree with that. The issue with software is that the people you make things for are usually anonymous and you'll never meet them, but if you've ever built software that helped someone and you witnessed it, it feels
38.
by suttontom 4mo ago
This is such a tired, meaningless argument. I've never seen a human in 10 years of professional software engineering at a large company ever so confidently, consistently create and send out seemingly well-reasoned code that's as w
39.
by suttontom 4mo ago
This is commonly known as "LLM-as-a-judge" and anecdotally multiple people I know who write code using OpenRouter or using multiple models say it's surprisingly effective. It's strange that there don't appear to be
40.
by suttontom 4mo ago
Ah yes, the magical equivalent of "you are a senior software engineer who writes bug-free code". IME people would benefit greatly from the process, albeit tedious and time-consuming, of testing out the same prompt sequence/se
41.
by suttontom 4mo ago
Isn't that kind of what they're doing with this rollout? Except they're just hand picking the companies.
42.
by suttontom 4mo ago
What is your problem? Do you think something is an opinion piece just because it has a byline? What about https://www.forrester.com/press-newsroom/forrester-impact-ai... ? Is there literally any evidence you'd acce
43.
by suttontom 4mo ago
Do you know if anyone has trained, say, a pre-2017 model and tried to get it to come up with Attention Is All You Need? If it did, would you say that was only because it's a synthesis of prior art? If so, what isn't?
44.
by suttontom 4mo ago
Are you joking? Is there literally "nothing" you can imagine that Claude can't do?
45.
by suttontom 4mo ago
https://www.shrm.org/topics-tools/news/technology/ai-layoffs... https://cmr.berkeley.edu/2025/10/seven-myths-about-ai-and-pr... https://www.technologyreview.com/2026
46.
by suttontom 4mo ago
This is a good example of being bad at writing code.
47.
by suttontom 4mo ago
Not to be cynical but do you think this would matter at all? Are you saying that companies would hold themselves to their missions or even something that's legally binding? > "Google is not a conventional company. We do not int
48.
by suttontom 4mo ago
Am I going crazy? Is a PR with 94 commits that adds 1,600 LoC actually considered "very reviewable"? Please someone tell me if I'm crazy?
49.
by suttontom 4mo ago
Models are not innately backwards-compatible. Both OpenAI and Anthropic encourage running evaluations and comparing the performance of your existing agent workflows against new models before just stepping up to the newest one because you ma
50.
by suttontom 4mo ago
I think LLMs are extremely useful, mostly for coding. But saying we're extremely close to an AI that can "reliably come up with novel actions for physical robots" feeds into the hype that these tools can do way or are very cl
51.
by suttontom 4mo ago
"UniSuper’s production Google Cloud VMware Engine (GCVE) private cloud was automatically deleted one year after it’s creation due to a misconfiguration in how it was created. When it was created, there was a bug in the creation script
52.
by suttontom 4mo ago
They can't even reliably follow instructions from text. I think "it's just around the corner/just wait x months/just wait and see bro" is one of the most telling signs of AI psychosis.
53.
by suttontom 4mo ago
You do know that this was the same thing people said about crypto, right? And that the internet of things where your fridge connects to the Internet is hated by most consumers and had nowhere near the impact that IoT evangelists said it wou
54.
by suttontom 4mo ago
This is such a creepy dystopian thing to say. Don't you realize that? Isn't this "yes, there will be pain, but the future is inevitable and we must go forward into it" attitude straight out of multiple horror and sci-fi
55.
by suttontom 4mo ago
Demos are also often misleading and cherry-picked. Using AI to do one cool demo that breaks down 99% of the time when circumstances slightly change has played an outsized part in most of the AI insanity we are living with.
56.
by suttontom 4mo ago
I'm amazed you think that instead of using an LLM that someone will go buy a book and spend a week learning something that, judging by the fact that they last used it 30 years ago, likely won't be relevant for them soon.
57.
by suttontom 4mo ago
I mean, yes, "sophisticated" people, including institutional investors, do fall for scams. See SBF, Theranos. These scams are often most effective when there's mania and FOMO about a technology that everyone seems to be sayin
58.
by suttontom 5mo ago
It's not even clear what this means. If an agent generates some code and I delete half and rewrite it, is the code AI-generated? If I start to write def calculate_average() and the "AI" auto complete fills it in when I hit ta
59.
by suttontom 5mo ago
It's really rare that you get someone claiming 10-50x productivity gains who posts some proof. It's not surprising that the person posting it has "contributed" nothing remotely of value except for a huge number of PRs th
60.
by suttontom 5mo ago
"The models will eventually..." Yeah but they haven't, and it's been years now. Also who cares? We have problems right now that need to be solved.

More ›