altdataseller

I disagree. No amount of fine tuning will ever give the LLM the relevant context with which to answer my question. Maybe if your context is a static Wikipedia or something that will never change, you can fine tune it. But if your data and docs keep changing, how is fine tuning going to be better than RAG?

For what its worth, i think its garbage as well and just wanted to post it here so more people could say how garbage it is and confirm my confirmation bias :)

Thats not the case: “ As of December 31, 2023, we had about 3,190 customers with ARR of $100,000 or more, an increase of 15% from 2,780 as of December 31, 2022.”

https://investors.datadoghq.com/news-releases/news-release-d...

Datadog grew revenue 25% from last year. So im not sure your narrative is true?

I doubt that will happen. One becsuse the leak didnt really disclose any major secrets that most marketers didnt already know.

2, even if some of this wasnt widely known, its not like you can take advantage of it overnight. Theres no quick hack to building a trustworthy domain or getting lots of trustworthy links for example

You can ensure you are crawled by simply not blocking them in your robots.txt. They'll find you eventually unless you simply never let anyone know about your website

Definitely not at that price point

My hair on fire problem right now is i want to get a list of all companies and their company descriptions, headcount range and sector that match a single company name, sorted by relevancy. If you can create an API for that and charge per API call (say $100 for 5000 calls), i would be your first customer

>> If you’re generating structured output (eg json, xml) which is not sequenced, and your output is something like {decision: …, reason:…} it literally does nothing.

Is this true if you are using RAG too?

Not doing market researvh for a competitor but more like want to get into the mind of these enterprise buyers and what they are thinking in general

Recommending kubernetes immediately is an insta-rejection for me, if I was an interviewer

Yes write SQL

Way way too complicated. I thought this tool was suppsed to make my life easier

Its not the early days in terms of expecting digital tools to be correct 99% of the time. Early adoption age was back in 2000-2009. Now everyone expects polished tools that does what it expects them to do

I thibk its the AI version of Clippy or something embedded in Office or Microsoft365 or whatever they call that thing these days.. (i think, i have absolutely no idea either :)

It should remind them their entire livelihood is dependent on one single company. They arent partners at all. Its more like one is sustenance/oxygen for the other

Ive used a less expensive plan years ago. The thing that bothered me the most was that they limited exports to 1000 or something unless you upgraded

Its not just a question of uptime but dependencies too. The moment it doesnt make strategic sense to have an open Bing API for Microsoft, anyone reliant on them is in for a world of hurt. Especially if its the core feature of your product (ie DDG)

Crunchbase has a data advantage because its parent conpany Techcrunch gets a lot of intel on startup funding.

>> If not, would you use one if I built it? I'm wondering how one might collect the data for just one category (to start)

If someone had a very good idea on how to collect that idea, they probably would be building it themselves….

Maybe this forces DDG and ChatGPT to make their own search engine index and corpus. Sure it might be a few years too late for the former, but thats probably what they said 5 years ago too.

None of those are essential.

The only essential thing it needs to do is store my files, with some assurity that they will exist x years from now in a cost efficient manner.

This is kinda meta but one problem i have is finding people in an industry to interview about their problems. you can of course spray and pray linkedin messages and hope someone is willijg to give you their time. Or ask your friends. But none of them are likely to be fruitful or it will take a long time.

What would help would be a network that will offer me 1 hour calls with profiessionals in an industry in exchange for $xxx

Ive said this in the other thread but come up with a problem first. You are putting the horse before the carriage or whatever. Come up wirh the privlem first. You cant worry about a gazillion thjngs at once

Come up with a problem first. The other stuff like whether it uses AI or whether some big corp will copy it is something you worry about later

Who are some noticeable examples?

I’m not a lawyer but does that mean scraping LinkedIn is indeed illegal? Or is it still yet to be determined?

Ironically, sometimes I wish that LinkedIn vs HiQ verdict went the other way and made scraping LinkedIn illegal

The only foolproof way is to have some proprietary dataset that OpenAI will never be able to get by themselves or from anyone else. That or be in a very small niche they will never bother to compete in because it’s not worth their time

Why is it useless?

I would guess the majority are high touch sales people reaching out to companies to advertise on the site, or managing their accounts