jacobr1

6 ms·

61.
by jacobr1 1y ago
An importantly the cycle time on this stuff can be much faster. Trying out different variants, and iterating through larger changes can be huge.
62.
by jacobr1 1y ago
Beyond code-gen I think some techniques are very underutilized. One can generate tests, generate docs, explain things line by line. Explicitly explaining alternative approaches and tradeoffs is helpful too. While, as with everything in this
63.
by jacobr1 1y ago
That provides a good case study. A single person can bootstrap a company, along with outsourcing and a ton of grit. Moving the boundaries on what tasks are automatable, at difference cost points, and different levels of experience required
64.
by jacobr1 1y ago
You _can_ build things that way. But then you need some business logic to decide which tools to expose to the system. The easy/dumb way is just to give it all the tools. With RAG, you have retrieval step where you have hardcoded some k
65.
by jacobr1 1y ago
You bought the bundle. You got games (crosswords and such), Local/National/Intl News, local Arts, schedule info for things like movies/theater, recipes, comics, sports score and analysis, classifieds, even for the ads to see
66.
by jacobr1 1y ago
The solution is to tax the land value rather than property improvement value. Any particular improvement from a property owner isn't taxed ... though implicitly the whole neighborhood getting more expensive (via the aggregate of all im
67.
by jacobr1 1y ago
Even in the case of the Milky Way blocking the light, we could probably do the math to estimate would an observer would see if they were outside the Milky Way, interpolating from the observed IR/UV and some model of the visible spectra
68.
by jacobr1 1y ago
Would not one approach be to just ensure the system has all the data? Relevance to address systems, side effects, and legal constraints. Then when making a recommendations it can account for all factors not just prior use cases.
69.
by jacobr1 1y ago
Not a answer, but contributory idea - Meta-analysis. There are plenty of strong meta-analysis out there and one of the things they tend to end up doing is weighing the methodological rigour of the papers along with the overlap they have to
70.
by jacobr1 1y ago
Yeah - that doesn't work. We include our own guidelines and practices in the context. Basically an ever evolving wiki page of PR best practices. We had that before trying LLMs - so it was easier for us to start. Also we found doing an
71.
by jacobr1 1y ago
Yep - this is the key to making it work
72.
by jacobr1 1y ago
Our team has pretty aggressively started using LLMs for automated code review. It will look at our PRs and post comments. We can adding more material for different things for it to consider- from a looking at a summarized version of our API
73.
by jacobr1 1y ago
The most cutting edge-models aren't usually free, at least at first.
74.
by jacobr1 1y ago
Seems the same to me. If you are using the llm as a tool to build your product (rather than a dependency within it for functionality) you can easily switch to a different model, or IDE/Agentic-coder in the same way you can switch betwe
75.
by jacobr1 1y ago
I've found LLMs to be a key tool in helping me articulate something clearly. I write down a few half-vague notes, maybe some hard rules, and my overall intent and ask it to articulate a spec, and then ask to for suggestions, feedback,
76.
by jacobr1 1y ago
Yeah, this example isn't great - you can just tell the llm to run pwd more frequently or something. But for the `$command | head -100` example, the usage is a bit different. I run into this myself on the cli, and often ended up using `
77.
by jacobr1 1y ago
Good point. Look at patents. Few are truly novel in some exotic sense of "the whole idea is something never seen before." Most likely it is a combination of known factors applied in a new way, or incremental development improving
78.
by jacobr1 1y ago
> You can't formally prove that some gene sequence is responsible for trait X etc. Maybe not formally in some kind of mathematical sense. But you certainly could have simulation models of protein synthesis, and maybe even higher ord
79.
by jacobr1 1y ago
It depends on the problem domain you have and the way you prompt things. Basically the reasoning is better, in cases where using the same model to critique itself in multiple turns would be better. With code, for example, if a single shot w
80.
by jacobr1 1y ago
from the parent post: `social media and/or ad data` So if you have ad impression data you have IP geolocation, or maybe better, along with the timestamp. Similarly for socials sometimes you get location metadata, and with image uploads
81.
by jacobr1 1y ago
Might be ok. I'm in the camp right now that you need run evals, test, and iterate to maximize your performance. But everything is going depend on the specifics of your use case - and that should be assessed empirically.
82.
by jacobr1 1y ago
The equivalent would be having the _same human_ review their own work. We require others with different experience and fresh eyes for secondary review and for the most important task multiple people. To some extent the same llm with a new c
83.
by jacobr1 1y ago
It is intended as a incremental step to now write revise the unsafe rust into more idiomatic and safe rust. I haven't done rust port yet, but when translating other projects between languages I've also found it useful to first do
84.
by jacobr1 1y ago
> And by the way, we are out of new training data to give the models. Only easily accessible text data. We haven't really started using video at scale yet for example. It looks like data for specific tasks goes really far too ... fo
85.
by jacobr1 1y ago
There seems to be lots of mixed data points on this, but to some extent there is knowledge encoded into the evolutionary base state of the new human brain. Probably not directly as knowledge, but "primed" to quickly to establish r
86.
by jacobr1 1y ago
It doesn't directly measure it, true. But the parent poster was saying it has extremely high positive correlation to measures that do measure it, so it probably doesn't matter.
87.
by jacobr1 1y ago
Right. Usually there is 100% authority in some kind of dictator. You instead have something that is more like an oligarchy. You have different interest groups with varying levels of influence. In some sense democracy works like this too. Ev
88.
by jacobr1 1y ago
Side consideration - my local utility has a program (targeted at lower income households - but there is no restriction on who can enroll) that average out the annual costs. For the first year they look at your total annual cost and divide b
89.
by jacobr1 1y ago
That is actually his step 1. Step 0 might be write down the problem as stated - but in his writing step 1 is really, "structure the problem." Of course you might have structured it wrong, that is in part where step 2 comes in.
90.
by jacobr1 1y ago
AI legible documentation. If you optimize for a "1-pager" doc you can add to the context of an LLM and that is all it needs to know to use your package or framework ... people will use it if has some kind non-technical advantage.

More ›