Live data from Hacker News

From where I left

antirez.com

21–30 of 472 posts

Re: From where I left

#21
post #15

Earlier quoted context omitted.

I covered this in the blog post: no big money involved, I asked to rejoin and without a big paycheck on purpose because to be free to act you should not get too much money, otherwise it's all fake.

But if you wanted to have influence without a big pay check, why not join Valkey? Redis, the company, in your absence has gotten a reputation as a predatory company that has attacked open source projects in other to further those goals. https://arstechnica.com/information-technology/2024/04/redis... https://news.ycombinator.com/item?id=39858144 https://news.ycombinator.com/item?id=42239607

Talking about predatory behavior https://www.forbes.com/sites/justinwarren/2024/04/11/opentof... https://www.ftc.gov/news-events/news/press-releases/2023/09/... https://apnews.com/article/google-antitrust-ad-tech-virginia...

Re: From where I left

#22
post #17
post #13

Earlier quoted context omitted.

Thank you Rui, my target is: API that can be learned and used under 120 seconds. I don't think more of this is needed, for the basic usage. Then there could be other stuff, but it must be incremental. Easy things must be easy.

Well, the VSIM example I got inside of 10 seconds, so at least for now you're on track :)

TLDR API:

VADD key [VALUES | FP32 ] element

VSIM key [VALUES ... | FP32 ... | ELE ...] [COUNT ] [EPSION ]

VCARD key (number of elements)

VDIM key (dimension of the vectors)

VDEL key

And so forth. However there are nice things you can enable, if you want. For instance VADD has the [REDUCE] option that implements random projection in order to reduce the dimensionality of your embeddings. Less memory, and it is faster. And so forth.

Will work on it during xmas, it's quite fun.

It implements on-insertion normalization and int8 quantization. The algorithm to insert nodes inside the HNSW attempts to never leave isolated nodes and to get good quality links and so forth. More hacking needed.

Re: From where I left

#23
> hint: (Claude AI is in its own league)

Is there a good way to use the Anthropic models as a private person? It's my understanding you can only pay for them as a business.

Re: From where I left

#24

I suspect he is going to make a lot of money coming back to save Redis from the negative press they have been having recently as well as the threat of Valkey. I sort of have to view it in light of all the other events that have been happening to Redis.

And Valkey with backing from AWS, GCP, and Oracle don't make any money at all ... wait aren't those trillion dollar monopolies? The pot calling the kettle black

Valkey also has the backing of The Linux Foundation. So maybe it just has a lot of backing in general.

https://www.linuxfoundation.org/press/linux-foundation-launc...

BTW are you an employee of Redis? You seem to have created this account just to reply to me.

Re: From where I left

#25
post #14
post #7

[flagged]

World does need unlimited amount of software and I think it can even be mathematically proven. Amount of software required grows exponentially because every piece of software then requires even more software on top of it - just like managing security issues in al the software built, etc.

That sounds just awful. Like an aggressive cancer that spreads until it kills the host and itself.

Re: From where I left

#26
post #23

> hint: (Claude AI is in its own league) Is there a good way to use the Anthropic models as a private person? It's my understanding you can only pay for them as a business.

You can use the API and Claude as a private individual. I've been doing so for a couple of years.

Edit: Go here and create an account: console.anthropic.com

Re: From where I left

#27
post #23

> hint: (Claude AI is in its own league) Is there a good way to use the Anthropic models as a private person? It's my understanding you can only pay for them as a business.

Nope it's like ChatGPT. 20 bucks / month.

Re: From where I left

#28
post #23

> hint: (Claude AI is in its own league) Is there a good way to use the Anthropic models as a private person? It's my understanding you can only pay for them as a business.

As sibling comments have said, you can pay per month to use the web interface, pay for use of the API via self-signup or you can even use them via openrouter.

Re: From where I left

#29
post #15

Earlier quoted context omitted.

I covered this in the blog post: no big money involved, I asked to rejoin and without a big paycheck on purpose because to be free to act you should not get too much money, otherwise it's all fake.

But if you wanted to have influence without a big pay check, why not join Valkey? Redis, the company, in your absence has gotten a reputation as a predatory company that has attacked open source projects in other to further those goals. https://arstechnica.com/information-technology/2024/04/redis... https://news.ycombinator.com/item?id=39858144 https://news.ycombinator.com/item?id=42239607

I want to help Redis to find a good path forward (while I respect the work of everybody). I was sponsored by this company for ages, I know the core hackers there, and they are strong and are good people. I talked with the CEO and he is very interested in finding novel ways to play, so I rejoined Redis and not ValKey.

Honestly even if Redis public image has not been stellar recently, I know the people that work at Redis, many engineers are there since the start, and I believe they have huge design abilities and can do what is needed to make Redis more useful for the community.

Re: From where I left

#30
post #22
post #17

Earlier quoted context omitted.

Well, the VSIM example I got inside of 10 seconds, so at least for now you're on track :)

TLDR API: VADD key [VALUES | FP32 ] element VSIM key [VALUES ... | FP32 ... | ELE ...] [COUNT ] [EPSION ] VCARD key (number of elements) VDIM key (dimension of the vectors) VDEL key And so forth. However there are nice things you can enable, if you want. For instance VADD has the [REDUCE] option that implements random projection in order to reduce the dimensionality of your embeddings. Less memory, and it is faster.…

Huh Charles actually can code? We’ll then.
Post reply on HN