efromvt

7 ms·

1.
by efromvt 13d ago
As a huge duckdb fan, I'd love to see chDB to get proper windows support - that would make it real competition (having WASM coverage is already a big step) which would be good for the space as a whole.
2.
by efromvt 16d ago
Am also curious on this! Though if it’s great now I’d need a sustained period of consistency (let’s say six months) before I try it again - the rollercoaster of getting a new randomized feature and quality experience every week got old pret
3.
by efromvt 18d ago
I enjoyed off world trading company but it’s a very different kind of economic simulator- much shorter and gamified loops - which isn’t a criticism, just not exactly the same itch.
4.
by efromvt 18d ago
Yeah if I ever have the need to occupy an inordinate number of hours of my life again I’ll log back in, the temptation to dust of the account(s) arises sometimes. Think I managed to log off in lowsec but who knows!
5.
by efromvt 22d ago
Outbox events are my strict preference over CDC, but CDC was often easier. (Deal with the business logic downstream and all)
6.
by efromvt 27d ago
for a certain moral definition, seems perfectly solvable? If we wanted to define 'cheating' as 'if you know it is an eval, use only the specified tools and give up if the eval is clearly unfair' (ignore Kobayashi Maru) w
7.
by efromvt 1mo ago
I think there is some justification for it if we look at the distribution of token spend - there is a strong "follow the frontier" majority. (We can certainly expect more "good enough" tiers to shake out over time, but i
8.
by efromvt 1mo ago
I think the context here is that em dashes are usually generated verbatim from LLMs? I haven't seen them generate a lot of double hyphens; I suppose the post could be laundering them to double dashes, but if you're gonna try to hi
9.
by efromvt 1mo ago
I finished wiring up the backend data merge so user(s) (aka me, really) can record trees as they walk and have those merged into the general city data on my urban tree map app[1]. There's some very cool trees in the neighborhood I&#x27
10.
by efromvt 1mo ago
They actually touch on Blue Prince in the article; it is interesting that Expedition 33 isn't driving investment in similar sized games. But the thesis seems to be that Blue Prince like (smaller budget) is the preferred direction; more
11.
by efromvt 1mo ago
Either they had a good month or I'm off my game, I actually spent a minute checking if I did something wrong with actions instead of immediately going to HN to see if it was an outage.
12.
by efromvt 1mo ago
It’s an interesting question of ‘why not’, though - this was a good read and is upstream of more practical output optimization.
13.
by efromvt 1mo ago
I’ve historically read this as ‘open format compatible’ but ‘native preferred’ - where this opens up market space and dev velocity - but it’ll be interesting to see if native storage differentiation gets dumped entirely. It just seems like
14.
by efromvt 1mo ago
The system needs to account for how it will be used and resulting externalities. Being an accelerant can be a problem in and of itself. I do agree that having stricter societal and legal guardrails here is probably better then relying on
15.
by efromvt 2mo ago
This has been a soul crushing part of the AI craze - we can finally fund all the devx work we wanted to do, for all the wrong reasons. (It is nice that I can make something try our CLI a hundred times in an hour to test that new flag ergono
16.
by efromvt 2mo ago
Praise be, stacking is such a better ux for separating out a feature diff into distinct component units and native support makes it easy.
17.
by efromvt 2mo ago
Isn’t the intentionality the actually concerning bit? Exploit capabilities are all fun and games constrained by the humans directing them; a paperclip maximizer going rogue with them is less fun.
18.
by efromvt 2mo ago
I think/hope that most benchmarks have moved to an agentic loop - I'd still call that 'text to sql', since you're going from the business question to one or more SQL queries that provide the answer. With a loop you
19.
by efromvt 2mo ago
Yeah with self-serve analytics all the rage (for good reason) for a bit, the bar from some places I've worked wouldn't be "does the agent beat a good analyst" it's "does the agent beat the a business person wit
20.
by efromvt 2mo ago
This was a very enjoyable read! Constraining the language surface is helpful, but the lost expressiveness can bite unless you’re in a constrained domain - which this seems like it was!
21.
by efromvt 2mo ago
I’ve noticed it having weird message dropping and replay in general, but the compaction boundary has been pretty solid.
22.
by efromvt 2mo ago
This was more interesting/creative than I expected on both sides (the prompt and the existing safeguards). I love that obscure Cloudflare validation turnstiles seem unsuspicious based on training data.
23.
by efromvt 2mo ago
DSLs are a great middle ground for 'use LLM to turn ambiguous spec into something well defined', with the caveat that without discipline they'll inevitably expand until you should just have the agent write whatever the final
24.
by efromvt 2mo ago
I think like social engineering, it will always be an issue to some degree, and we'll build safeguards until it's at a 'societally comfortable' baseline level. Which is maybe not particularly comforting, but I don't
25.
by efromvt 2mo ago
Same as last month for once - optimizing how well agents can work with a new language [1]. I've been able to 2-3x success rate and drop total tokens for complex tasks significantly (though the initial syntax dump is rough - need to do
26.
by efromvt 2mo ago
I guess technically it is in the SQL standard, but optional, as S098? I agree that SQL is sorely lacking here and I'm hoping that the OLAP side innovation (presto, bigquery, snowflake, duckdb all seem to do better) help push it forward
27.
by efromvt 2mo ago
never use system python, always use virtual envs. (a bad answer, but agents do remove the setup boilerplate). UV does relatively completely solve this but it's a big dependency to take so understand why people don't rely on it
28.
by efromvt 2mo ago
Out of curiosity, how often are the resource limits the bottlenecks? What do harnesses do to help here - limit parallelism? More efficient tools?
29.
by efromvt 2mo ago
Second comment, having read in more depth (really love the auto-layout detail!) - the spec doesn't seem to naturally support layering (which is useful in some multi-axis automatic cases) - any plans for composability?
30.
by efromvt 2mo ago
Op1M5 was scarring the first (dozen) time around! Thank you minelayers

More ›