throwaway7783

4 ms·

1.
by throwaway7783 7d ago
I mean, that goes without saying for LLMs. It is an approximation of human knowledge after all.
2.
by throwaway7783 7d ago
LLMs are becoming a world expert in everything. I am exploring this exact area of search and analytics for vanilla postgres as replicas. Guess what, the LLM came up with this exact conclusion of using ctids as docids, all by itself. It was
3.
by throwaway7783 7d ago
Is the middle option asking LLM to generate a classic ML model? Or generate tons of them and pick the best?
4.
by throwaway7783 12d ago
Cap the number, sort by salaries (perhaps by coarse domains of skills and not very narrow . Nurses vs Doctors. But s/w engineers are s/w engineers). Testing abilities is a gamable proxy. How much a company is willing to spend on
5.
by throwaway7783 12d ago
They cannot yet outmatch humans in critical thinking. Heck, they cannot outmatch a mediocre person like me yet. How many times have we awen "You are right...." Even on sota models? They are great tools for research and tasks as of
6.
by throwaway7783 13d ago
Do you think Flock is in the same category of installing PCs for police departments, when it comes to public impact?
7.
by throwaway7783 14d ago
In the short term yes. Long term, no. But I guess you can say the same thing about a team of engineers taking shortcuts
8.
by throwaway7783 14d ago
All the whistleblowers have said the alignment projects are underfunded. Surely they don't need external help to fund that? Could've done years ago?
9.
by throwaway7783 14d ago
The cynicism is about motivations and not that they are inherently not bad. Perhaps they are as bad as they claim. Or perhaps they're worse. All we have a couple of run of the mill breach examples and some people inside the talking abo
10.
by throwaway7783 14d ago
Yep. And the difference is clear as da for anyone using them both. And in spite of that advantage, OAI is now trying out ads. I can only imagine that even they are getting constrained to compute and are trying to find other ways to plug it
11.
by throwaway7783 14d ago
Ban non US models and form a cabal, with the blessings of the government. That's what it is looking like, no?
12.
by throwaway7783 14d ago
Same can be said about a hundred other things in the world. All the way from knives to nuclear.
13.
by throwaway7783 14d ago
When corporations are involved, it is always a good bet to err towards cynisim. From my own standpoint, Claude has started sucking really bad (incoherent, uncontrollable verbosity slow and so on) and I stopped using it. OpenAI started expe
14.
by throwaway7783 17d ago
Yes. This one is a major issue. It keeps track of various changes it made in the same session and tries to be backward compatible. Have to repeatedly tell not to be backward compatible. Things that are automatic for humans and aren't e
15.
by throwaway7783 23d ago
Well, I prefer the polars version. And now if I want to reuse the CTEs elsewhere, I have to reach out to another tool like DBT or hand roll something to do string manipulation.
16.
by throwaway7783 23d ago
My issue with DBT is it is a mix of SQL, yaml, jinja2 flow controls (and metrics is whole another thing). SQL with jinja2 if/else can get really unmaintainable quickly. It's perhaps better than homegrown sql based transformers. po
17.
by throwaway7783 23d ago
we have pre-commit hooks for fast tests (under a minute on a 2024 mac), a playwright script (smoke tests) that can be run on-demand locally, including setup and cleanup (~5 minutes). And on top of this we also have CI tests pre-merge (which
18.
by throwaway7783 27d ago
Maybe using git notes for the conversation or conversation summary is better.
19.
by throwaway7783 1mo ago
My go-to has been replicas for each use case, with well defined semantics for replication lags. I'm working on something that does most of this seamlessly (transactional,search, columnar & time series, vectors and queues) without h
20.
by throwaway7783 1mo ago
It is not just the software. Let's not forget the existing customer base, an opportunity cost to build something as mature as openrouter (even with LLMs). Perhaps it is worth it for them.
21.
by throwaway7783 1mo ago
We keep reinventing everything. Eclipse style plugins from a long time back
22.
by throwaway7783 1mo ago
Also a very nice attack surface.
23.
by throwaway7783 2mo ago
This is a great project. Thank you! A question on 20s postgresql time - It does not look like you are accounting for reading data from disk? Wouldn't the aggregation query have to load data from disk first? Or is it somewhat guaranteed
24.
by throwaway7783 2mo ago
Is this the Galleria test?
25.
by throwaway7783 2mo ago
Absolutely. I have been so livid at so many applications for not providing a decent CDC API (and dont forget deletes). Salesforce perhaps is the best out there. Imagine if every application exposed a standard CDC API, the world of integrati
26.
by throwaway7783 2mo ago
While it is frustrating, I do think disability is a bug (I have some). Just like I take medicines to manage my condition, or use glasses to read better (I don't ask everything to be made bigger - the text on pages in a book - I use an
27.
by throwaway7783 2mo ago
Isn't this the direction Claude on the web is going towards? It's a mix of message view on the left and specific interactive UI on the right panel. There is also MCP Apps spec now
28.
by throwaway7783 2mo ago
That second point is almost never true " the same REST/data API can serve your website and other consumers, e.g. any native mobile apps or third party API consumers" In a reasonably complex application, each interface ends up
29.
by throwaway7783 2mo ago
Yes, but very few people are on that edge of product and technology. From that angle LLMs are destroying the junior engineer dev market
30.
by throwaway7783 2mo ago
I agree on this one. Junior developers are not yet good guiding hands yet. But for the motivated junior developers LLMs are a great onboarding and learning tool. A good senior engineer must still guide, and there are very large gaps on how

More ›