Hacker Newsnew | past | comments | ask | show | jobs | submit | early_exit's commentslogin

Good read! How many pages of text were in scope? I'm not sure if the 1615+1629 pages were the total or just a subagent.

If they were the total I would say it was arguably more impressive the author was able to narrow it down to just 3000 pages than it was to find the dodo mention amongst those!


Those are years, not page counts, right? The article mentions "millions of records", but I'm not sure how big a record can be.


As far as I can tell, Opus 5.5 just did a roundabout grep for the Portuguese word for dodos.


for me it basically all boils down to:

1. I dont want to have to accept every time Claude touches our DB

2. I'm scared out of my mind it might do something bad to the DB

Plan mode gives me enough confidence that it wont do (2) --> allowing me to give it enough permissions to do (1)


Oh gods, I don't give it write access to my actual DB.

For my small-scale sqlite dB, it gets read access, and I encourage it to test modifications by copying it somewhere and writing into that.

Scale-dependent, but I hope to not have to work at a scale where it gets write access to the production DB. That just seems like asking for trouble.


FYI I had Clod attempt to corrupt a prod db the other day. (Opus 5)

I was experimenting with a rather complicated backfill operation, were I had a validation script I understand and have Clod come up with the backfill script. I was running against a local prod copy, and it proposed running the actual (unfinished) backfill script against prod.

It didn't have access to the secrets and I also caught the command, but a good reminder that this stuff needs guardrails.


yeah that's crazy. its so good 99% of the time but I've seen it have some insane hallucinations before (as late as Fable... cant remember if it was Fable 5 or Fable 5.1).

Hallucination not a big deal when it's on the surface layer. But I can't imagine the damage it could do if it hallucinated while building/validating a "load-bearing" component and then continued down that path


Make a db replica or just a db user account with readonly permissions, and have only those in your env, or docs accessible to agent. It's liberating.


Do you mean when you're making changes to a production DB?


yeah (at an early stage data-focused startup). I think when our product is a bit more mature we'd probably have a staging tier with full access and then manually promote builds / DB changes to prod. So many things to build lol

we DO daily snapshotting, so the risk is limited... but still spooks me


> data-focused startup

What could possibly go wrong with building a data-focused company on a foundation of violating the most fundamental precepts of data management


Write schema/data migration files [with agents] in version control and deploy those deterministically (after review).


I'm not so convinced on the land rent thesis.

Fewer jobs, less money --> people less willing to pay for anything more than bare minimum --> downward pressure where being cheap is the biggest selling point.


The thesis (pre-AI) is not "fewer jobs, less money."

It's "more productivity, more money → higher land rents → upward pressure on everything → higher floor of economic productivity required for someone to earn a plot of land to live or work upon"


To be fair, "them" are actively working to replace "us" with AI.


As a young Have Not, I'm just waiting this one out. I see 1 of 2 options in the future:

1. Prices go down because nobody can buy the homes at their current prices given current interest rates

2. Prices stay up because large buyers scoop up the homes. Over time, fewer and fewer Americans actually own homes. Eventually the majority of voters will be Have Nots and the scales of power will tip.


It's not mega investors who are buying most homes. Most homes are just going to people that saved a fuckload of money over the pandemic or people that already have a home and can use its equity to get another one. The economy is doing very well, I don't think there will be any shortage of people who want homes, even at current interest rates and prices.


History says that #2 is where we are heading. That was Piketty's main thesis in Capital in the 21st Century. The era of relatively inequality we saw after WW2 was an aberattion. We are returning to the historical norm of high-inequality.

Still remains to be seen how much closer we have to get to feudalism before anything happens to change our trajectory.


Given Twitters current valuation, wouldn’t Musk consider any outcome where he pays less than 5B an absolute win?


Matt Levine posted an outcome that seems pretty plausible.

Twitter settles and buys Musk's shares for $0 worth ~$2.6b at current price.

But it would stop Musk strategically dumping the stock and so could be worth much higher to Twitter.


I presume Musk would consider paying 0 an "absolute win". I kind of doubt he'll achieve that though.


Do these compare performance before or after fees? Regardless, if you're comparing funds (that contain lower risk investments, like bonds) to the S&P500 then they're going to lose whenever equity markets beat bonds - which is most of the time.

If I actually (by picking stocks) have a lower expected return than the S&P, shouldn't I be sending all the hedge funds my stock picks so that they can short them (while being long S&P) - therefore beating the benchmark ~75% of the time?


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

Search: