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

ChatGPT for Teams

openai.com

Post reply on HN
Print view View original

161–170 of 453 posts

Re: ChatGPT for Teams

#161

Post by toyg »

defrost wrote: post #103
benreesman wrote: post #86

Earlier quoted context omitted.

On reflection this thread is pretty clearly of general interest and my comment was more jargon that language, I hang out in ML zones too much. For a broad introduction to the field Karpathy's YouTube series is about as good as it gets. If you've got a pretty solid grasp of attention architectures and want a lively overview of stuff that's gone from secret to a huge deal recently I like this treatment as a light but p…

It was entertaining to follow and cross check for sanity - admittedly I first read "RenTech" as a reference to https://theitcrowd.fandom.com/wiki/Reynholm_Industries .. which seemed to fit suprisingly well.

Ironically, one of the best scenes from the series (Reynholm trying to wake his desktop by shouting "HELLO COMPUTER! HELLO!!" for hours and hours) feels very "ai"...

Re: ChatGPT for Teams

#162

Post by lysecret »

shrx wrote: post #142

Could a moderator change the "Teams" in the title to lowercase (as it is in the article)? Capitalizing Teams misleadingly implies it's regarding Microsoft's chat platform.

Ok, big noob question I never understood how to capitalize headlines correctly. Can someone eli5 ?:D

Re: ChatGPT for Teams

#163

Post by callalex »

anonylizard wrote: post #69
kannangce wrote: post #65

I thought teams would be cheaper than individual.

Have you ever seen an enterprise plan cheaper than the individual plans (Which are often free)? Now normal software is priced to squeeze as much money as you can, enterprises can afford more, so are charged more. Individuals are highly price sensitive, so has to be very cheap. GenAI is quite different in that its not 0 marginal cost, the marginal costs are probably at least 50% of the price. So the price difference b…

No, per-seat cost for these kinds of things is ALWAYS cheaper than retail. However it does require you to set up a meeting with a sales rep who will then work tirelessly to expand the number of services you use and require longer commitments etc. With “enterprise” pricing, the sticker price is just the opening number in a negotiation, and basic theory tells us that the opening number must be large since it sets a ceiling.

Re: ChatGPT for Teams

#164

Post by benreesman »

transitus wrote: post #140
Demiurge wrote: post #74

Earlier quoted context omitted.

I have 20 years of software development experience, and I couldn’t understand anything you said. Is there a dictionary for this new lingo, or am I just too mid?

We're all too mid. Luckily, these days we hoomans have AIs to help us understand other hoomans. Here is Gpt-4-1106-preview and Perplexity.ai versions trying to shed some light what was being said. https://pastebin.com/JuxfdrLg Hilariously neither knows who is sama (Sam Altman, the Drama King of OpenAI), nor do they recognize when they themselves are being discussed. Reading the responses in full also gives you a glim…

I absolutely love pointing these things at each other and watching them go.

I screen-capped my take on this to prove* that I was actually wiring all this stuff up and plug my nascent passion/oss project, but it's really funny comparing them either way: https://imgur.com/WDrqxsz

Re: ChatGPT for Teams

#165

Post by pests »

Roark66 wrote: post #104
alwa wrote: post #55

Earlier quoted context omitted.

For some time I thought the primary way to opt out was to disable chat history. That made it impossible to use custom instructions. But I later found the account-level opt-out request, which, assuming they respect the weakly-worded “ask us not to train on your data” language, preserves the product’s full set of features: https://privacy.openai.com/policies

There is literally nothing on that page (nor in the linked help article that describes how to export data) that says how to tell them to "not use my data for training" and not loose history at the least. If you know more about this feature please do tell.

?

Roark66 gave a better answer, but...

On desktop I clicked the link and immediately saw "Make A Privacy Request" top right (where login / account / menu buttons might be)

I must ask did you honestly just miss this UI element or do you think it might be some confirmation bias that you already had?

Re: ChatGPT for Teams

#166

Post by pests »

cald0s wrote: post #157
shrx wrote: post #142

Could a moderator change the "Teams" in the title to lowercase (as it is in the article)? Capitalizing Teams misleadingly implies it's regarding Microsoft's chat platform.

But that's how headlines work

I agree, but why didn't the original article do it? Or did they see the potential for confusion and deliberately not?

When a common word is a product / brand, how do you use that word in a title without bringing up associations with that product / brand?

Isn't the guidelines to submit the title as-is, no editorializing?

Re: ChatGPT for Teams

#167

Post by rvense »

lysecret wrote: post #162
shrx wrote: post #142

Could a moderator change the "Teams" in the title to lowercase (as it is in the article)? Capitalizing Teams misleadingly implies it's regarding Microsoft's chat platform.

Ok, big noob question I never understood how to capitalize headlines correctly. Can someone eli5 ?:D

In English, the most common convention is to capitalize every word, or every word except small ones like articles and prepositions (the, of, for).

Re: ChatGPT for Teams

#168

Post by theshackleford »

wokwokwok wrote: post #118
MacsHeadroom wrote: post #94

Earlier quoted context omitted.

Mistral Medium beats 4.5 on the censorship benchmark. It doesn't refuse to help with anything that could be vaguely non-PC or could potentially be used to hurt anyone in the wrong hands, including dangerously hot salsa recipes.

That's not a metric. That's a use case. Certainly, no one here is arguing that there are things openai refuses to allow, and given that the effectiveness of using GPT4 on them is literally zero , a sweet potato connected to a spring and keyboard will "beat" GPT-4, if that's your scoring metric. If you want a meaningful comparison you need tasks that both tools are capable of doing , and then see how effective they ar…

> If you want a meaningful comparison you need tasks that both tools are capable of doing, and then see how effective they are.

The fact it’s incapable of simple requests that an alternative can is absolutely part of a worthwhile comparison.

Re: ChatGPT for Teams

#169

Post by EmilStenstrom »

benreesman wrote: post #66

I’ve got my stuff rigged to hit mixtral-8x7, and dolphin locally, and 3.5-turbo, and the 4-series preview all with easy comparison in emacs and stuff, and in fairness the 4.5-preview is starting to show some edge on 8x7 that had been a toss-up even two weeks ago. I’m still on the mistral-medium waiting list. Until I realized Perplexity will give you a decent amount of Mistral Medium for free through their partnership…

Here's a glossary to understand this post:

- mixtral-8x7 or 8x7: Open source model by Mistral AI.

- Dolphin: An uncensored version of the mistral model

- 3.5-turbo: GPT-3.5 Turbo, the cheapest API from OpenAI

- 4-series preview OR "4.5 preview": GPT-4 Turbo, the most capable API from OpenAI

- mistral-medium: A new model by Mistral AI that they are only serving through AI. It's in private beta and there's a waiting list to access it.

- Perplexity: A new search engine that is challenging Google by applying LLM to search

- Sama: Sam Altman, CEO of OpenAI

- RenTech: Renaissance Technologies, a secretive hedge fund known for delivering impressive returns improving on the work of others

- DPO: Direct Preference Optimization. It is a technique that leverages AI feedback to optimize the performance of smaller, open-source models like Zephyr-7B1.

- Alibi: a Python library that provides tools for machine learning model inspection and interpretation2. It can be used to explain the predictions of any black-box model, including LLMs.

- Sliding window: a type of attention mechanism introduced by Mistral-7B3. It is used to support longer sequences in LLMs.

- Modern mixtures: The process of using multiple models together, like "mixtral" is a mixture of several mistral models.

- TheBloke: Open source developer that is very quick at quantizing all new models that come out

- Quantize: Decreasing memory requirements of a new model by decreasing the precision of weights, typically with just minor performance degradation.

- 4070 Super: NVIDIA 4070 Super, new graphics card announced just a week ago

- MSFT: Microsoft

Re: ChatGPT for Teams

#170

Post by bambax »

benreesman wrote: post #136
ChatGTP wrote: post #85

Earlier quoted context omitted.

Same bro

My apologies to both you and grandparent, I faux-pass'd on using such sloppy jargon in a thread of general interest. If I can be useful in decoding some of that and helping to keep this topic accessible to the broader community I'd be happy to answer any questions about what I meant. I've cited a few useful resources elsewhere in the thread and am always eager to talk shop on topics like this, so fire away if you hav…

What is "Mixtral Medium"? Is it a next version of Mixtral and can it be downloaded somewhere?
Post reply on HN