Hacker News new | past | comments | ask | show | jobs | submit | rbranson's comments login

Did you read the article? Dziri and Peng are not the “skeptical AI community,” they are in fact die hard AI researchers. This is like saying people who run benchmarks to find performance problems in code are skeptics or haters.

I read the article: it does not look like very good research: It's simple to find flaws in LLMs reasoning / compositional capabilities looking at problems that are at the limit of what they can do now, or just picking problems that are very far from their computational model, or submitting riddles. But there is no good analysis of the limitations, nor inspection of how/how much better recently LLMs got exactly at this kind of problems. Also the article is full of uninformative and obvious things to show how LLMs fail in stupid tasks such as multiplication between large numbers.

But the most absurd thing is that the paper looks at computational complexity in terms of direct function composition, and there is no reason an LLM should just use this kind of model when emitting many tokens. Note that even when CoT is not explicit, the LLM output that starts to shape the thinking process still makes it able to have technically unbound layers. With CoT this is even more obvious.

Basically there is no bridge between their restricted model and an LLM.


The issue is that law enforcement is flagrantly committing serious crimes, for their own profit. It’s offensive and disgusting.


Would they be able to profit if these items weren't scarce?


tl;dr is that "ultra-processed foods" are engineered to be more palatable and that tends to result in more caloric density.


One of the only (the only?) commercial grade implementations was launched recently by us at PlanetScale:

https://planetscale.com/blog/announcing-planetscale-vectors-...


No ability to host offline, and for 1/8th CPU + 1GB RAM + 800 GB storage, the price is $1,224/month?

I'm sure it works great, but at that price point, I'm stuck with self-hosting Postgres+pgvector.


Just pointing out that what you're paying for is actually 3x these resources. By default you get a primary server and two replicas with whatever specification you choose. This is primarily for data durability, but you can also send queries to your replicas.


Which works completely fine as long as you know how to manage your own db without getting wrecked!

But yes, I it seems extereme. But it is also cheaper than hiring a dedicated postgres/db guy who will cost 5 to 10x more per month.


There are plenty of set-it-and-forget-it vector dbs right now, maybe too many![0]

[0]https://news.ycombinator.com/item?id=41985176


For sure, I personally use pgvector myself but I also don't have millions and millions of rows. I haven't messed with anything other than Pinecone so I can't speak to those services, but there's a big difference than a vector db for your own personal use and a chat app/search on a db with millions of users convos and docs, not sure how well these managed vector DB platforms scale, but you probably need the db guy anyways when you're using vectors at scale. Atleast I would.


What's the advantage of NN over vectordb anymore? Are we losing some info when we embed?


It works great. We’ve had SPANN in production since October of 2023 at https://turbopuffer.com/


SPANN is also implemented in the open-source Vespa.ai


Actual SPANN or janky "inspired by SPANN" IVF with HNSW in front? Only real SPANN (with SPTAG, and partitioning designed to work with SPTAG) delivers good results. A superficial read of the paper LOOKS like you can achieve similar results by throwing off the shelf components at it, but it doesn't actually work well.


This is a really long blog post to just say that you should turn on the WAL if you want concurrency out of SQLite. All the other stuff is superfluous.


Can't agree.

I learned about BEGIN IMMEDIATE TRANSACTION.

And there's also busy_timeout.

The article also explains why/how/when things occur in detail which is valuable.


WAL mode makes those redundant. Just use WAL mode.


We all wish you were right. But alas life's not that simple.

I suggest reading the manual on the section: "Sometimes Queries Return SQLITE_BUSY In WAL Mode"

https://sqlite.org/wal.html#sometimes_queries_return_sqlite_...


Yeah, this almost never happens in practice. It’s not even worth being concerned about.


Happened all the time to me before I did some tuning Depends how much write contention you have and how long open transactions take to finish.


The list of circumstances for WAL-mode busy errors is in the doc linked by one of the posters above. It has nothing to do with transactions.


That list lists reasons queries might return busy; queries aka. reads.

Reads returning busy is rare under WAL, but WAL mode does very little for writer-writer contention.


…what? No it doesn’t. Go read the article. Every optimization listed addresses a different performance aspect of SQLite.


How does busy_timeout address a performance aspect of SQLite?


As I explain in the post, if you have multiple connections and consistent write load, the timeout will penalize older quereres and noticeably harm your long tail latency


I thought it was a pretty good list of common Rails-application-specific and sqlite3-specific knobs to turn, for newcomers to performance tuning. (Really just a guided tour though -- turn this knob to enable this particular tool for dealing with concurrency problems...)


Yeah calling the EC2 API is definitely more complex than leasing datacenter space, purchasing racks of hardware, deploying a fault tolerant and secure network, capturing and managing offsite backups, dealing with hardware component failures, etc.


Serious question, are you familiar with dedicated hosting?


If you’re at that scale, there are plenty of other platforms that don’t have vendor lock in and overly complex, proprietary APIs by design.



I was under the impression we were talking about ICs here -- your link shows that an upper-middle IC that you'd expect to be choosing between early startup or FAANG will see something around 450 a year, which tracks much closer to what I'd expect.


You said 1 or 2 people who aren't executives. There's way more than 1-2 E7s at Meta, as an example.


That's how risk works. The FAANG employee friends have exactly a 0% chance at a 9-figure outcome. They'll easily be at top decile if they pull a nominal $10M+ post-tax in 20 years.


The fact that New York (a city of 8M people) has only received ~150% more funding than Mountain View (a city of 81K people) tells you all you need to know about the promise here.


Mountain View is a tiny part of the Bay Area, which just so happens a population of ~8M. So, Apples/Apples maybe?

Or should we talk about AI startups in East Harlem only?


The promise that New York doesn't hinge on the existence of a single industry?


Mountain View is part of a contiguous suburb stretching between South San Francisco, San Francisco, and San Jose which have massive biotech, tech, and manufacturing industries, respectively.

It doesn't hinge on any one of them.


New York is part of a contiguous megalopolis stretching between Boston and Washington DC, which features the same industries.


Oh oh, I sense a return of the 'East vs West Computer Bowl' coming. Someone get Stewart Cheifet on the line!



Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: