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

> Even worse, well over half of the SVB bailout went to <20 accounts.

Do you have a source for this?


I don't have time to research now, but there were some well detailed breakdowns of the large payments when the bailout happened.

The first relevant link in DDG is a statement by the Chairman of the FDIC. He claims that the top ten accounts held more than 13.3B between them, which is more than 10% of the total of all deposits in the bank, and more than half of the $20B that the government is expected to payout in total. [1][2]

The FDIC would otherwise have had to pay a scant 3M of that, so there's 2/3 of the bailout just in those 10 accounts. If we include the next ten accounts, it will be much, much more. (One of the articles at the time claimed more than 85% went to 15 accounts, but I didn't want to be extreme without time to research sources.)

[1]: https://www.banking.senate.gov/imo/media/doc/Gruenberg%20Tes... [2]: https://www.fdic.gov/news/press-releases/2023/pr23023.html


I think that's a misinterpretation of the data. Based on reports, SVB had around 130B of deposits when it was taken into receivership. Something like 90% of that was uninsured, so around 115B of uninsured deposits. The 20B hole means that without FDIC backing the uninsured deposits, they still could have paid out over 80 cents on the dollar. So if the top ten accounts had about 13.3B of deposits, they only received less than 2.7B of the bailout - less than 15%, not "well over half".


Accepted, and thank you for the clarification.

At the time of the collapse there were definitely articles making that claim, with analysis and numbers, and I still assume it to be correct.

When I saw this, I was surprised, it seemed too clean and straightforward. Will have to look for the original sources.


I strongly suspect that any sources that claimed that simply made the same error that you did.

Looking at the source you provided from the Chairman of the FDIC, it actually says that the top ten accounts held $13.3B - not _more than_ $13.3B, as you claimed. The next ten accounts necessarily held less than that, so say a generous upper bound at $26.6B for the top twenty accounts. In order for them to get at least half of the bailout, they would have needed to hold north of $55B in aggregate. The FDIC itself would have to have gotten the size of SVB's largest depositors wrong by over double in order for your original claim to be true, in the absolute most generous case.


Heya - sign up for an account and put in your personal access token, it will work then! Hitting GitHub rate limits atm!


It uses the OpenAI embedding models to get around the context window problem. You can compress large amounts of text/code into just a few tokens.


Can you elaborate on this? What is the input to the GPT in that case? I was under the impression that GPT is given an array of tokens and it produces one token as its output.


Would love some more information on how to do this. Can you share details or point to a resource?



I'm curious if can share LoC -> token count metrics?


Hey guys - creator here. Have hit the Github rate limit and currently not working! Am trying to fix! Thanks for all the support!


Might be worth asking users to "login with GitHub" to get past the rate limits (the anonymous rate limits are lowwww like 10 per hour).

Logging with GitHub is still lowish but like 1k per hour or so. (IIRC)


I didn't expect this many people to try it out haha - with hindsight, absolutely should have


There’s no way I’m logging into rando apps using my GitHub id. Aside from their bad UX and the potential to accidentally over grant permissions, the app could abuse GitHub ToS with my credentials and get me banned.

Not worth the risk just to play around with a site.


Funny that github is the bottleneck rather than the OpenAI for once.


Do the request from the browser and that won't happen because the request source will be distributed.


Need to download the files to the server to generate the embeddings!


Update: just shipped a fix for the rate limits - you will need to signup for an account and put your own personal access token in.


The sign-up form doesn't seem to allow passwords with non-alphanumeric characters in them. Ironically, when it sees one, it complains that "password must have at least one number and at least one letter" (even though it already does) - I suspect that's just a catch-all message for invalid passwords?

In any case, the restriction makes it incompatible with pretty much all password generators, including the built-in ones in modern browsers. To improve matters, I would suggest dropping all restrictions on what cannot be included altogether.


Had a broken validator - fixed!


shakes fist at Microsoft


How are you avoiding hitting the rate limit for open ai apis



fastly RCA is underwhelming - no info on what was the component, what happened and how the situation was tackled.


I think a public company will never dare go into as much detail as bunny did. Or maybe it is just the size of the organisation that discourages that.


Cloudflare's RFO blog posts are incredibly detailed. Each time I read one, I feel reasonably confident that they have learnt from the mistakes that lead to that outage and that it shouldn't happen again.

https://blog.cloudflare.com/tag/outage/


Yeah, but Cloudflare is accountable to the US Congress, they better be transparent.


TreeCard | Mobile and Backend Engineers | Full-Time or Contract to Hire | Remote

TreeCard is a debit card which reforests the planet when you spend, backed by Ecosia. We have over 150,000 people on our waiting list, we’ve just closed a $5mn seed, and our ambition is to be the leading green finance brand for the 21st century. By 2026, we aim to have planted 1bn trees.

Hiring for mobile (React Naive) and backend (Go) engineers, please reach out at gary AT treecard.org


FluidStack | Golang Engineer | P2P Networking Engineer, Infrastructure Engineer | London | ONSITE | Full-time | http://www.fluidstack.io

FluidStack has built technology to turn any device into a cloud server, creating a massively distributed cloud platform running on consumer devices. We leverage under-utilised capacity to provide lower cost and higher performance cloud services. Long term, we want to create a distributed AWS.

We are VC-backed (Episode 1, Seedcamp, Founders Factory), and are seeking world-class engineers to help us scale the platform as we work towards our Series A at the end of the year. Drop me an email on gary [AT] fluidstack.io if you are interested in learning more!


FluidStack | Peer to Peer (P2P) Networking Engineer, Infrastructure Engineer | London | ONSITE | Full-time | http://www.fluidstack.io

FluidStack is building technology to turn any device into a cloud server, creating a massively distributed cloud platform, cheaper and faster than incumbents.

We are VC-backed (Episode 1, Seedcamp, Founders Factory), and are seeking world-class engineers to help us build out and scale the platform. More detailed job details on our site.

Drop me an email on gary [AT] fluidstack.io if you are interested in learning more!


Flare | Peer to Peer (P2P) Networking Engineer, WebRTC Engineer, Cloud Systems Engineer | London | ONISTE | Full-time | http://www.flare-global.com

Flare is building technology to turn any device into a cloud server. We aim to use this to build a cheaper, faster, and more ethical internet.

We are VC-backed (Episode 1, Seedcamp, Founders Factory), and have built out a working first version. We are now seeking world-class engineers to help us build out and scale the platform.

Our jobs are here: https://angel.co/flare-10/jobs Drop us an email at team [AT] flare-global.com


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

Search: