whakim

7 ms·

1.
by whakim 4d ago
Sure; total blinding is an issue with many drugs (however active placebos which can make you feel dizzy/sedated/whatever can help); opiods/fentanyl/etc sometimes have to make do with imperfect blinding. The fact remains
2.
by whakim 5d ago
> The "war on drugs" certainly didn't help, but research never really stopped. This is false. The scheduling of psychedelics such as psilocybin and LSD was catastrophic for medical research in the United States (as well as
3.
by whakim 5d ago
There are a number of problems with these statements, but a major one is that psychedelics are associated with reduced activity within the DMN (but increased activity between the DMN and other networks). It is not clear the direction in w
4.
by whakim 2mo ago
I agree that there's plenty of footguns with async python, but I think the hype is a bit overblown. A few sensible defaults in the standard library would go a long way, e.g. requiring a maxsize argument to Queue (or None if you really
5.
by whakim 3mo ago
> But as the environment of the organism gets more complex, reflexive avoidance behavior isn't sufficient for competence. For an agent in a complex environment, competent damage avoidance requires engaging with negative valence as a
6.
by whakim 3mo ago
> For example, phenomenal pain brings with it competence at protecting bodily integrity. The memory of pain becomes part of the explanatory narrative for the monitoring function that tracks progress towards goals ensuring coherent behavi
7.
by whakim 3mo ago
Sure. We can't prove that other organisms experience qualia; we can only look at the effects of qualia (e.g. behaviors that are likely to be the product of emotions) and assume that an organism is therefore conscious. The real point,
8.
by whakim 3mo ago
> The recognition of oneself as situated in the world is crucial to coherent engagement with the world. It is how an entity can ensure its body parts are moving towards the same goal. It's how behavior over time doesn't undermi
9.
by whakim 3mo ago
Isn't Maturana's theory that consciousness has to do with language, and the use of language to make distinctions about ourselves and others? To me, this seems clearly insufficient to explain consciousness - qualia totally precede
10.
by whakim 4mo ago
What a wild take. The straightforward takeaway from the end of ZIRP and the resulting increase in focus would be that you need to say no to more things, not fewer? You have to really contort yourself to argue that actually ZIRP gave rise
11.
Turning RAG pipelines into enterprise-grade Data Subscriptions(halcyon.io)
6 points by whakim 5mo ago|0 comments
12.
by whakim 6mo ago
I don't think we should undersell that transformers and semantic search are really powerful information retrieval tools, and they are extremely potent for solving search problems. That being said, I think I agree with you that RAG is f
13.
by whakim 6mo ago
I'd argue the author missed a trick here by using a fancy embedding model without any re-ranking. One of the benefits of a re-ranker (or even a series of re-rankers!) is that you can embed your documents using a really small and cheap
14.
by whakim 6mo ago
For technical domains, stuffing the context full of related-and-irrelevant or possibly-conflicting information will lead to poor results. The examples of long-context retrieval like finding a fact in a book really aren't representative
15.
by whakim 6mo ago
There is no reason to believe that the foods humans have historically eaten are safer/healthier than "industrially processed/extracted/refined" food simply because we have historically eaten them. Evolution does not
16.
by whakim 7mo ago
Yes; you can phone it in post-tenure. But just because it is possible doesn't mean (in my experience) it is common; and I don't think it's helpful (as TFA claims) to equate this possibility with "a total scam." To g
17.
by whakim 9mo ago
Worth noting that the filtering implementation is quite restrictive if you want to avoid post-filtering: filters must be expressible as discrete smallints (ruling out continuous variables like timestamps or high cardinality filters like ids
18.
by whakim 9mo ago
> Most of the time you don't need a different Python version from the system one. Except for literally anytime you’re collaborating with anyone, ever? I can’t even begin to imagine working on a project where folks just use whatever
19.
by whakim 9mo ago
I don't really understand the point around error handling. Sure, with structured outputs you need to be explicit about what errors you're handling and how you're handling them. But if you ask the model to return pure text, yo
20.
by whakim 10mo ago
In my experience the semantic/lexical search problem is better understood as a precision/recall tradeoff. Lexical search (along with boolean operators, exact phrase matching, etc.) has very high precision at the expense of lower r
21.
by whakim 10mo ago
Doesn't this depend on your data to a large extent? In a very dense graph "far" results (in terms of the effort spent searching) that match the filters might actually be quite similar?
22.
by whakim 11mo ago
Thanks for the reply! This makes much more sense now. To preface, I think pgvector is incredibly awesome software, and I have to give huge kudos to the folks working on it. Super cool. That being said, I do think the author isn't being
23.
by whakim 11mo ago
> maintenance_work_mem begs to differ. HNSW indices are big . Let's suppose I have an HNSW index which fits in a few hundred gigabytes of memory, or perhaps a few terabytes. How do I reasonably rebuild this using maintenance_work_m
24.
by whakim 11mo ago
Interested to hear more about your experience here. At Halcyon, we have trillions of embeddings and found Postgres to be unsuitable at several orders of magnitude less than we currently have. On the iterative scan side, how do you prevent t
25.
by whakim 1y ago
Ok, but what was the cost of labor put into curation of the training dataset and performing the fine-tuning? Hasn’t the paper’s conclusion been repeatedly demonstrated - that it is possible to get really good task-specific performance out
26.
by whakim 1y ago
No offense, but this software engineering elitism does no favors to perceptions of the field. In reality, most other fields are complex and the phenomenon of believing something is simple because you don't understand it is widespread a
27.
by whakim 1y ago
You enforce them; if I get a ticket for parking at an intersection, I won't do it again! (Also, in the specific context of this discussion, parking restrictions near intersections are super common; this is not some esoteric new law t
28.
by whakim 1y ago
No, I don't; there are plenty of places you can't legally park that do not have painted curbs or "No Parking" signage. Do we also need curbs and signage near every fire hydrant? How about every driveway? Can drivers doub
29.
by whakim 1y ago
Why? Having a driver's license is a privilege that requires you to study and know the rules of the road. The onus is on you to know the rules.
30.
by whakim 1y ago
I couldn't agree with this more. I don't think the majority of problems with vector search at scale are vector search problems (although filtering + ANN is definitely interesting), they're search-problems-at-scale problems.

More ›