Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ex-Googlers raise $40M to democratize natural-language AI

fastcompany.com

Post reply on HN
Print view View original

91–100 of 126 posts

Re: Ex-Googlers raise $40M to democratize natural-language AI

#91

Post by h0mie »

Tenoke wrote: post #87
___luigi wrote: post #81

Earlier quoted context omitted.

Yes. HF is built on a tiered subscription (SaaS) model. https://huggingface.co/pricing You can refer to this discussion for more details: https://twitter.com/migueldeicaza/status/1285204129225281536

I believe huggingface mostly subsist on investments and the amount they earn directly is still pretty small compared to expenses, is that no longer the case?

One of the founders said on twitter (don't remember when exactly, but less than a year I think) thay they were cash flow positive.

Re: Ex-Googlers raise $40M to democratize natural-language AI

#92

Post by tgv »

dna_polymerase wrote: post #71
rahimnathwani wrote: post #28

Earlier quoted context omitted.

What they say they're doing fits one of the definitions of democratize: "make (something) accessible to everyone". More about the etymology here: https://www.etymonline.com/word/democratize

All that source tells us is that the word has been wrongly used since the french revolution. Democracy means governance by the masses. Has nothing to do with making things accessible to the masses.

Les docteurs du spin...

Re: Ex-Googlers raise $40M to democratize natural-language AI

#93

Post by streetcat1 »

___luigi wrote: post #68

In the ML industry, there are two type of companies. 1/ Open source companies like huggingface ( https://github.com/huggingface ), explosion ( https://github.com/explosion ), Fast.ai that democratized access to ML and provided a set of tools & language models for engineers. These companies didn't only manage to build a company and open source projects, they built a welcoming community and it's impressive to see all t…

Unfortunately the new models are very big, hence they cannot be released as smaller models (as open source). They must be served as an api.

Re: Ex-Googlers raise $40M to democratize natural-language AI

#94

Post by jasode »

abeppu wrote: post #8

Which is more "democratized": a large language model which can be downloaded, and accessed as a library, originally based from the work of the FAANG giants (e.g. huggingface transformers) or an API, where every invocation is a call that flows through Cohere's servers?

>Which is more "democratized": a large language model which can be downloaded, and accessed as a library, [...] or an API, where every invocation is a call that flows through Cohere's servers?

I do understand the point you're trying to make: local autonomy is superior to cloud access mediated by a private commercial entity.

However at this time, we may have a counterintuitive situation where API access is more "democratic" than downloading a huge model.

Based on various reports[1], the GPT-3 model was trained on ~45 terabytes of text corpus (Wikipedia + Web Common Crawl + book texts, etc) and the final runtime model (175 billion parameters) requires ~350 gigabytes of RAM. In that case, the model size is ~1% of the training set.

So "democractize" depends on how ambitious the user is. If you want to use a very large 350GB RAM model, the cloud model with API will be more accessible by the masses than running on local hardware. Last time I looked, an Intel Xeon motherboard has max ram of 128GB so scaling up to 350GB RAM is not going to be cheap or trivial to build.

Let's further extrapolate to a future hypothetical GPT-4 using ~10x multiplier: train on 450 terabytes of text with a model requiring 3.5 terabytes of RAM. How do we make that future huge model accessible to the masses? Probably via a cloud API. Unfortunately, there's an unavoidable hardware capital expense barrier there.

[1] https://www.google.com/search?q=gpt-3+350gb+ram

Re: Ex-Googlers raise $40M to democratize natural-language AI

#95

Post by barbecue_sauce »

drdeca wrote: post #77
NhanH wrote: post #66

Earlier quoted context omitted.

So from your definition and example, it means that Coke want to democratize soda, Apple want to democratize iPhone and Tesla want to democratize EV. I don't think those sentence would have your expected meaning when someone else read them.

Well, no, because the current state of affairs isn’t that very few people have access to smartphones / soda .

So Coke pioneered the democratization of soda.

Re: Ex-Googlers raise $40M to democratize natural-language AI

#96

Post by occamrazor »

jasode wrote: post #94
abeppu wrote: post #8

Which is more "democratized": a large language model which can be downloaded, and accessed as a library, originally based from the work of the FAANG giants (e.g. huggingface transformers) or an API, where every invocation is a call that flows through Cohere's servers?

>Which is more "democratized": a large language model which can be downloaded, and accessed as a library, [...] or an API, where every invocation is a call that flows through Cohere's servers? I do understand the point you're trying to make: local autonomy is superior to cloud access mediated by a private commercial entity. However at this time, we may have a counterintuitive situation where API access is more "democ…

If OpenAI made the weights and the NN configuration freely available,I am sure that several organizations would offer cheap (or even free with strict rate limits) API access. And users would also be able to run the model locally, if they can afford the hardware.

Re: Ex-Googlers raise $40M to democratize natural-language AI

#99

Post by WanderPanda »

jasode wrote: post #94
abeppu wrote: post #8

Which is more "democratized": a large language model which can be downloaded, and accessed as a library, originally based from the work of the FAANG giants (e.g. huggingface transformers) or an API, where every invocation is a call that flows through Cohere's servers?

>Which is more "democratized": a large language model which can be downloaded, and accessed as a library, [...] or an API, where every invocation is a call that flows through Cohere's servers? I do understand the point you're trying to make: local autonomy is superior to cloud access mediated by a private commercial entity. However at this time, we may have a counterintuitive situation where API access is more "democ…

This is a strawman argument. Publishing the code/weights is not mutually exclusive to providing an API

Re: Ex-Googlers raise $40M to democratize natural-language AI

#100

Post by WanderPanda »

beaconstudios wrote: post #82
abeppu wrote: post #8

Which is more "democratized": a large language model which can be downloaded, and accessed as a library, originally based from the work of the FAANG giants (e.g. huggingface transformers) or an API, where every invocation is a call that flows through Cohere's servers?

When a startup says anything, you should hear "we plan to use this to get hella rich". There is no such thing as a startup with altruistic intent.

But it is also not zero sum. Long term viable companies provide a higher value to their customers than the price they charge.
Post reply on HN