At this point their products have so little material that it may be better for the environment to buy an IKEA one vs decent furniture (from the standpoint of caring for the environment)
Very editorialised title. Not sure this offers any real value to be honest other than to drum up hype before IPO valuations hit for some of these companies.
I think the impact is heavily exaggerated. Yes, it means mathematics will be transformed into a field that relies on tools instead of your mind alone. Just like most fields. Doctors heavily rely on (really expensive!) tools to make diagnoses and do surgeries too. But it doesn't make them less valuable. If anything it makes them more valuable, as the society needs a long training and filtering process to decide who to use these tools.
The math being simple while the effects emerge automatically is the best part. Relativity is unintuitive precisely because the consequences are so far from the formula.
I do not believe you read my comment in good conscience.
>>Claude was consistently doing database operations in a for loop, where it could use batch operations.
>>The code in the end is better than a lot of the code I've seen humans write.
>It's a little amusing to see those two sentences written back to back with no hint of irony to be frank.
There is no irony because I have seen humans do the same thing. The difference is that the human was paid 50x more.
I do not understand your arguments around outsource shops. I get clients. They want something. I deliver. There is not a lot of complexity to this. I am not a world start programmer that has 50 clients to choose from.
>>Set up excellent testing.
>Is this the part that you hand code or constantly review yourself? I guess not, since you would have explicitly mentioned something that important. I would caution you not to be surprised when no one believes you have excellent testing when you've unleashed multiple LLMs on it and are not reviewing code anymore.
I set up excellent testing the same way a CTO would set up excellent testing: I give clear guidelines on the architecture I want to see on the project. I then review key parts and samples of it and give feedback. When it looks good I mark it as done and move on. Any bugs or defects always result in additional test coverage.
I'm happy to have a constructive discussion about this, but I do not appreciate your sarcastic, tongue-in-cheek, condescending tone. It makes people not wanting to communicate with you.
Schema changes need locking... everywhere. Citus is no different. And they need proper design everywhere. If you mean that it requires distributed transactions, well, yes, again: expected and solved. Not even all sharding solutions support this.
> coordinator node
Not sure what the problem is here. If what you mean is that a single coordinator, even with an HA replica, can saturate, that's true, but you can add multiple "query routers" (that's our name in StackGres, see [1]).
> Maybe I am wrong, but I am yet to read stories on operating tens of TB scale workloads on citus.
For example, we have a customer that ingests some 30TB/day, and it's ramping up towards 200TB/day of ingestion. On 24 worker nodes.
This tracks with my experience. I have assembled a homebrew microbiology lab from various parts from aliexpress and integrated it into my agent harness. Just a year ago it was a complete mess, but current models are extremely good at synthesising novel protein and virus structures. One put me in the hospital not last week, haha! I cant wait to see what I perish of
Honestly, maybe it's a market specific thing, but from having worked for a carrier, the overwhelming majority of consumers just dont know anything about any phone but the brand. They'll come in store looking for "an iPhone" or "a Galaxy", and the employee will usually push them towards whichever phone has the best deal (and very often, incentives to be sold), which rarely happens to be these alternative models like the Mini. The minority who does buy their phones outright or know what they want does tend to gravitate towards larger phones, but the rest typically just buys the standard, base model.
I'm pretty sure the AI companies could train an abort feature into the LLM, but they have no incentive to do so.
Having LLMs break out of sandboxing is free marketing for them and it reduces the amount of resources spent on things that don't improve benchmark results.
This is from the intellectual con artist Terence McKenna. McKenna never let things like this get in the way of his theory.
There was a flaw in the math of his ridiculous Time Wave Zero idea that when pointed out to him, didn't change anything.
I was super into McKenna as a dumb teenager. I even had the dos program of Time Wave Zero.
The idea McKenna would be taken seriously on anything after December 21, 2012 is really an interesting study in the power of the cult of personality.
The stoned ape idea has more holes than swiss cheese.
None of the least being that the actual study he based this idea on with edge detection didn't even say what he claimed it did. Distorted form perception as opposed improved edge detection, something that makes perfect sense to anyone who has tripped.
He is not even a clever fraudster. It is just hard to know as a dumb teenager that James Joyce and Whitehead's Process and Reality are intentional choices in the unintelligible and the obscure so they could mean whatever McKenna needed them to mean.
'21st-Century' in the title is doing the work of distinguishing, not describing. An open-licensed textbook with homework is the actual novelty, and that's worth putting in the title.
Loom is such an amazing project. I was always surprised by the fact that similar tooling for C/C++ is much more cumbersome to use, since that's where the memory model came from.
>Not a Lean expert but some of the proof tactics used to prove are probably novel
Maybe. But they're in 32 millions lines of Lean. How do you find the needle in the heystack ?
>Or, you could prompt agents later to analyze which lemmas or parts of the proof are surprising or applicable to other problems?
If OpenAI was truly serious about improving maths (and not jerking themselves off), they'd have also used Prove2Me (and contributed their results back), which would have done that. Each part of the proof combines into a larger graph, that everyone can reuse. Note that Anthropic isn't better there: yes, they used Prove2Me, but as far as I know they haven't contributed back to it, and just shat out 10 million lines and a good luck everyone.
Spider webs and auto traffic also have a relationship in this sense but that doesn't say much about the relationship of cars and spiders.
Maybe you need to read less bullshit from McKenna. He was a con artist, not even a good one at that.