mediaman

11 ms·

151.
by mediaman 1y ago
Simon’s content is not spam. Spam’s primary purpose is commercial conversion rather than communicating information. Your goal seems to be discourage people from writing about, and sharing, their thoughts about technical subjects. To whateve
152.
by mediaman 1y ago
And a topk of 30 also means reranking of any sort is out, except for maybe limited reranking of 30->10, but that seems kind of pointless with today’s LLMs that can handle a bit more context.
153.
by mediaman 1y ago
This is one reason why verifiable rewards works really well, if it's possible for a given domain. Figuring out how to extract signal and verify it for an RL loop will be very popular for a lot of niche fields.
154.
by mediaman 1y ago
Interesting theory, but Amazon uses tons of stochastic methods (including deep reinforcement learning) throughout the business, including warehouse inventory management. "Determinism" is not some north star that operations people
155.
by mediaman 1y ago
The website the human sees is the new API. That's needed because many APIs are either nonexistent or extremely marginal in design and content coverage.
156.
by mediaman 1y ago
It's interesting to see software engineers realize that AI can be useful in the hands of competent engineers, but that LLMs tend to produce a mess in the hands of those with little software engineering knowledge. Then they're conf
157.
by mediaman 1y ago
New supply obviously increases demand because shifting the supply curve to the right (more quantity) reduces the clearing price, which increases demand. That's econ 101. In this classic case, increasing supply both decreases price and
158.
by mediaman 1y ago
Yes, people are going to get squeezed out of the housing market by new arrivals if new housing stock isn't built. "Squeezing out" is done by a price mechanism: a family that would prefer to stay in Seattle decides to sell, be
159.
by mediaman 1y ago
This analogy seems confused. If someone in Chicago moves to Seattle, then our policy options are (1) no new condo in Seattle or (2) new condo in Seattle. Under policy 1, the new buyer from Chicago must outbid locals for the fixed housing su
160.
by mediaman 1y ago
Keyword search (or something similar in concept, like bm25) would typically be first stage, rather than second, since it can be done with an inverted index.
161.
by mediaman 1y ago
I actually don't agree. Tool use is the key to successful enterprise product integration and they have done some very good work here. This is much more important to commercialization than, for example, creative writing quality (which i
162.
by mediaman 1y ago
It works on Codex CLI, install it with npm. That's been out for a while and used their 'codex' model, but they updated it today to default to gpt-5 instead.
163.
by mediaman 1y ago
This is endlessly brought up as if the human operating the tool is an idiot. I agree that if the user is incompetent, cannot learn, and cannot learn to use a tool, then they're going to make a lot of mistakes from using GPTs. Yes, ther
164.
by mediaman 1y ago
He's not saying it's only true when salaries are equal. It makes a significant difference at the margin: if you're an employer debating whether to employ US talent or offshore talent, weighing time zones, skill and other fact
165.
by mediaman 1y ago
It's a tool. It's not a human. A line number works great for humans. Today, they're terrible for LLMs. I can choose to use a screwdriver to hammer in a nail and complain about how useless screwdrivers are. Or I can realize wh
166.
by mediaman 1y ago
AI is less useful when I'm picking through specific, careful business logic. But we still have lots of boilerplate code that's required, and it's really good at that. If I need a reasonably good-looking front-end UI, Claude w
167.
by mediaman 1y ago
What do you recommend for this? I've actually had good luck having them create XML, even though you're "supposed" to use the native tool calling in a JSON schema. There seems to be far fewer issues with getting JSON synt
168.
by mediaman 1y ago
The original story doesn't make any sense. How would you fake an "AI" agent coding by using people on the other side? Woudn't it be...obvious? People cannot type code that fast. What's your non-snarky theory about h
169.
by mediaman 1y ago
You're bringing up the question of "what is a fair tax rate," which is reasonable. The questions of this legislation, though, are different: Should we incentivize companies to hire corporate executives instead of engineers? S
170.
by mediaman 1y ago
NOLs are generally not transferrable in the US (they used to be, but now the benefit can only be used if the acquirer of the 'rump' continues the existing operating business).
171.
by mediaman 1y ago
Salary, mostly. It's useful to separate out the GPU cost of training from the salary cost of the people who design the training systems. They are expensive. That does not mean, however, that inference is unprofitable. The unit economic
172.
by mediaman 1y ago
A lot of people don't know what this Section 174 is about, so here's a brief explainer. Normally, when you have expenses, you deduct them off your revenue to find your taxable profit. If you have $1 million in sales, and $900k in
173.
by mediaman 1y ago
That's the old way of thinking about software economics, where marginal cost is zero. Marginal cost of LLMs is not zero. I come from manufacturing and find this kind of attitude bizarre among some software professionals. In manufacturi
174.
by mediaman 1y ago
What are you doing to get results this bad? I tried this question three times and each time the first two products met both requirements. Are you doing the classic thing of using the free version to complain about the competent version?
175.
by mediaman 1y ago
Just like some people who wrote long sentences into Google in 2000 and complained it was a fad. Meanwhile the rest of the world learned how to use it. We have a choice. Ignore the tool or learn to use it. (There was lots of dumb hype then,
176.
by mediaman 1y ago
If an LLM improves coding productivity, and it is better at one language than another, then at the margin it will affect which language you may choose. At the margin means that both languages, or frameworks or whatever, are reasonably a
177.
by mediaman 1y ago
That only works if there are moats that discourage new entrants (e.g., buying a social media upstart works because of the network effect). Not obvious serverless postgres is in that category.
178.
by mediaman 1y ago
There are no more “old school” production lines left? I’m not sure what you mean; obviously there are a lot of production lines in the US, including the manufacturing of engine blocks as shown here. Do you mean that they use robots today an
179.
by mediaman 1y ago
The cotton gin processed short fiber cotton, but not long fiber cotton. Did the cotton gin therefore not compete with human labor?
180.
by mediaman 1y ago
That’s funny, because if you go into any machine shop you’ll find machinists writing things down in notebooks (and probably cursing engineers who never set foot in the shop).

More ›