ChatGPT Enterprise
openai.com
openai.com
From our discussions with enterprises (trying to sell our LLM apps platform), we quickly learned how sensitive enterprises are when it comes to sharing their data. In many of these organizations, employees are already pasting a lot of sensitive data into ChatGPT unless access to ChatGPT itself is restricted. We know a few companies that ended up deploying chatbot-ui with Azure's OpenAI offering since Azure claims to not use user's data (https://learn.microsoft.com/en-us/legal/cognitive-services/o...).
We ended up adding support for Azure's OpenAI offering to our platform as well as open-source our engine to support on-prem deployments (LLMStack - https://github.com/trypromptly/LLMStack) to deal with the privacy concerns these enterprises have.
You’ve got to be early, but not so early you get legal or business disruptions or concequences.
It’s quite the balancing act for exec teams.
So why do we care from where LLMs learn?
Because humans aren't computers and the similarities between the two, other than the overuse of the word "learning" in the computer's case, are nonexistant?
There is a legal difference between learning from something and truly making your own version and simply copying.
It's vague of course - take plagiarism in a university science essay - the student has no original data and very likely no original thought - but still there is a difference between simply copying a textbook and writing it in your own words.
Bottom line - how do we know the output of the LLM isn't a verbatim copy of something with the license stripped off?
same difference there is between painting your own fake Caravaggio and buying a fake Caravaggio (or selling the one you made).
the second one is forgery, the first one is not.
Because humans dont put the "Shutterstock" watermark logo on the images they produce.
There are some startups working in the space that essentially plan to do something like this. https://www.konfer.ai/aritificial-intelligence-trust-managem... is one I know of that is trying to solve this. They enable these foundation model providers to maintain an inventory of training sources so they can easily deal with coming regulations etc.
Microsoft/OpenAI are selling a service. They’re both reputable companies. If it turns out that they are reselling stolen data, are you really liable for purchasing it?
If you buy something that fell of a truck, then you are liable for purchasing stolen goods. But if it turns out that all the bananas in wall mart were stolen from cosco you’re not as a customer liable for theft.
Similarly, I don’t know if Clarkson Intelligence have purchased proper license for all the data they are reselling. Maybe they are also scraping some proprietary source and now you are using someone else’s IP.
Actually, that would be fencing stolen goods and customers could have obligations.
Cases of bananas is a bit silly as returning bananas would be not possible and value is too small to bother.
But imagine reputable car dealer selling stolen cars, repossession is far more likely here.
How legitimate the IP concern is and whether it holds up in court is one thing, but finger pointing will probably not be sufficient.
Realistically you can prove that just as well as you can prove that employees aren't using ChatGPT via their cellphones.
There are also organizations that forbid the use of Stack overflow. As long as employees don't feel like you're holding back their career and skills by prohibiting them from using modern tools, and keep working there, hey. As long as you pay them enough to stay, people will put up with a lot, even if it hurts them.
Official documentation is still available…
To effectively sue you, I believe the plaintiff would have to prove the LLM you were using was trained on that IP and it was not in the public domain. Neither seems very doable.
Ex. CoPilot can insert whole blocks of code with comments and variable names from copyrighted code, if those aspects are sufficiently unique then it's extremely unlikely to be produced any way other than coming from your code. If the code isn't a perfect copy then it's trickier, but that's also the case if I copy your code and remove all the comments, so it's still not all that different from the current status quo.
The bigger question is who gets sued, but I can't imagine any AI company actually making claims about the copyright status of the output of their AI, so it's probably on you for using it.
Being able to have your legal counsel tell them to go bug openAI could potentially save you from quite a few anklebiters all seeking to get their own piece.
From my view, copying information from Google search results isn't that much different from copying the response from ChatGPT.
Notably Stack Overflow's license is Creative Commons Attribution-ShareAlike, which I believe very people actually realize when copying snippets from there.
A lot of the snippets would not meet the standard for copyrightable code, though. At least that’s my understanding as non-lawyer.
We won't know if this is legally sound until a company who isn't forbidding A.I. usage gets sued and they claim this as a defense. For all we know the court could determine that, as long as the content isn't directly regurgitated, it's seen as fair use of the input data.
i.e. Without ChatGPT an employee could still copy and paste something from somewhere. ChatGPT actually doesn't change the equation at all.
We also provide support and some premium processors for enterprise on-prem deployments.
Except many companies deal with data of other companies, and these companies do not allow the sharing of data.
The SOC2 framework is complex and compliance can be expensive. This can lead organizations to focus on ticking the boxes rather than implementing meaningful security controls.
SOC2 is not a good universal metric for understanding an organization's security culture. It's frightening that this is the best we have for now.
I personally believe that in addition to OpenAI's offering, the ability to swap to an open source model e.g. Llama-2 is the way to go for enterprise offerings in order to get full control.
"They're huge pussies when it comes to security" - Jan the Man[0]
[0] https://memes.getyarn.io/yarn-clip/b3fc68bb-5b53-456d-aec5-4...
They'll want to climb the compliance ladder to be considered in more highly-regulated industries. I don't think they're quite HIPAA-compliant yet. The next thing after that is probably in-transit geofencing, so the hardware used by an institution reside in a particular jurisdiction. This stuff seems boring but it's an easy way to scale the addressable market.
Though at this point, they are probably simply supply-limited. Just serving the first wave will keep their capacity at a maximum.
(I do wonder if they'll start offering batch services that can run when the enterprise employees are sleeping...)
OpenAI offers baa to select customers.
You can get a BAA through Azure’s OpenAI service though, I believe the details are located in this document:
https://azure.microsoft.com/en-us/resources/microsoft-azure-...
https://help.openai.com/en/articles/5722486-how-your-data-is...
That said, for enterprises that use the consumer product internally, it would make sense to pay to opt-out from that input being used.
What is actually stopping them? Most companies won't have the fire power to go against microsoft backed openai. How can we ensure that they can't violate this? How can they be practically held accountable?
This as far as I am concerned is "Trust me bro!". How is it not otherwise?
Are you claiming this because they used copyrighted material as training data? If so, I think you're starting from the wrong point.
Please correct me if I'm wrong, but last I heard using copyrighted data is pretty murky waters legally and they're operating in a gray area. Additionally, I don't think many open source licenses explicitly forbid using their code as training data. The issue isn't just that most other companies don't have the resources to go up against Microsoft/OpenAI, it's that even if they did, it isn't clear whether the courts would find that Microsoft/OpenAI did anything wrong.
I'm not saying that I side with Microsoft/OpenAI in this debate, but I just don't think this is as clear cut as you're making it seem.
All open source license comes under copyright law. It means if they violate the OSS license, the license is void and the tech/material becomes copyright protected. So yes, it would mean that it is trained on copyrighted material.
> Additionally, I don't think many open source licenses explicitly forbid using their code as training data.
It doesn't forbid. For example, permissive license like MIT can be used to train LLM's if they are in compliance. The only requirement when you train on a MIT licensed codebase is that you need to provide attribution. It is one of the easiest license to comply. It means, you just need to copy paste the copyright notice. The below is the MIT license of Emberjs.
Copyright (c) 2011 Yehuda Katz, Tom Dale and Ember.js contributors
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
This copyright notice needs to be somewhere in ChatGPT's website/product to be in compliance with MIT license. If it is not, MIT license is void and you are violating the license. The end result is you are training on copyrighted material. I am more than happy to be corrected if you could find me any single OSS license attribution shown somewhere for training the openai model.
Also, this can be still be fixed by adding the attribution for the code that is trained on. THIS IS MY ARGUMENT. The absolute ignorance and arrogance is their motivation and agenda.
Which is why I am asking, WHAT IS STOPPING THEM FROM VIOLATING THEIR OWN TERMS AND CONDITIONS FOR CHATGPT ENTERPRISE?
First offense could be excused as "blazing a trail and burning down the forest by accident".
But now they have a direct business contract with bigger companies that can lawyer up way better than open source foundations that live on donations and goodwill of code contributors.
Imagine they make a huge deal with Sony or Dell and either company can prove their "secure" enterprise plan was used for corporate espionage.
The legal and reputation repercussion could sink even a fortune 100 company
The same would apply to any other application. What if company A uses code from company B via ChatGPT/CodePilot because company B's code was used as training data? Imagine a startup database company using Oracle's database code through use of this technology.
And if a proprietary company accidentally uses GPL code through these tools, and the GPL project can prove that use, then the proprietary company will be forced to open source their entire application.
Top 1 misconception about open source licenses.
GPL doesn't mean if you use the code your entire project will become GPL.
GPL means if you use the code and your project is not GPL-compatible, you are committing copyright infringement. As if you stole proprietary code. If brought to the court, it would be resolved just like other copyright infringement cases.
Microsoft/Amazon/Google already have competitor's data in their cloud. They could even fake encryption to get all the customer's disk access. Also most employees use google workspace or office 365 cloud to store and share confidential files. How is different with OpenAI that makes it any more worrying?
> For all enterprise customers, it offers:
> Customer prompts and company data are not used for training OpenAI models.
> Unlimited access to advanced data analysis (formerly known as Code Interpreter)
> 32k token context windows for 4x longer inputs, files, or follow-ups
" We do not train on your business data or conversations, and our models don’t learn from your usage. ChatGPT Enterprise is also SOC 2 compliant and all conversations are encrypted in transit and at rest. "
That's great. But can customer prompts and company data be resold to data brokers?
ChatGPT Enterprise is also SOC 2 compliant and all conversations are encrypted in transit and at rest. Our new admin console lets you manage team members easily and offers domain verification, SSO, and usage insights, allowing for large-scale deployment into enterprise.
I think this will have a solid product-market-fit. The product (ChatGPT) was ready but not enterprise. Now it is. They will get a lot of sales leads.
Such dark patterns, plus their involvement in crypto, their shoddy treatment of paying users, their security incidents... make it harder for me to feel good about OpenAI spearheading the introduction of (real) AI into the world today.
That's great. But can customer prompts and company data be resold to data brokers?
- GPT-4 (ChatGPT Plus): has max 4K tokens ?
- GPT-4 API: has max 8K tokens (for most users atm)
- GPT-3.5 API: has max 16K tokens
I'd consider the 32K GPT-4 context the most valuable feature. In my opinion OpenAI shouldn't discriminate in favor of large enterprises. It should be equaly available to normal (paying) customers.
Having conversations saved to go back to like in the default setting on Pro, that's disabled when a Pro user turns on the privacy setting, is another big difference.
This is borderline extortion, and it's hilarious to witness as someone who doesn't have a dog in this fight.
Companies are trying to maximize profit; they are not trying to minimize costs so they can continue to do you favors.
These arguments creep up frequently on HN: "This company is doing X to their customers to offset their costs." No, they are a company, and they are trying to make money.
I don't think they're removing all instances of your company from their existing data sources, which would make sense to call "borderline extortion".
Fine-tuning isn't great at learning knowledge. It's good at adopting tone or format. For example, a chirpy helper bot, or a bot that outputs specifically formatted JSON.
I also doubt they're going to have a great system for fine-tuning. Successful fine-tuning requires some thought into what the data looks like (bare docs won't work), at which point you have technical people working on the project anyway.
Their future connection system will probably be in the format of API prompts to request data from an enterprise system using their existing function fine-tuning feature. They tried this already with plugins, and they didn't work very well. Maybe they'll come up with a better system. Generally this works better if you write your own simple API for it to interface with which does a lot of the heavy lifting to interface with the actual enterprise systems, so the AI doesn't output garbled API requests so much.
- Codebases
- Documents (by way of connection to your Box/SharePoint/GSuite account)
- Knowledgebases (I'm thinking of something like a Notion here)
I'm really looking forward to seeing what they come up with here, as I think this is a truly killer use case that will push LLMs into mainstream enterprise usage. My company uses Notion and has an enormous amount of information on there. If I could ask it things like "Which customer is integrated with tool X" (we keep a record of this on the customer page in Notion) and get a correct response, that would be immensely helpful to me. Similar with connecting a support person to a knowledgebase of answers that becomes incredibly easy to search.
They support uploading documents to it for that via that code interpreter, and they're adding connectors to applications where the documents live, not sure what more you're expecting.
Edit: spelling
That being said you can use fine tuning to improve retrieval, which indirectly improves recall. You can do things like fine tune the model you're getting embeddings from, fine tune the LLM to craft queries that better match a domain specific format, etc.
It won't replace the expensive on-the-fly retrieval but it will let you be more accurate in your replies.
Also retrieval can be infinitely faster than inference depending on the domain. In well defined domains you can run old school full text search and leverage the LLMs skill at crafting well thought out queries. In that case that runs at the speed of your I/O.
TLDR: This might have just killed a LOT of startups
Personally I love what Evenup Law is doing. Basically find a segment of the market that runs like small businesses and that has a lot of repetitive tasks they have to do themselves and go to them. Though I can't really think of other segments like this :)
If you want to build something uniquely useful, you probably have to do your own training at least.
If you want to build an AI start-up and need a LLM, you must use Llama or another model than you can control and host yourself, anything else is basically suicide.
It's not free if you have paying clients.
> If you want to build an AI start-up and need a LLM, you must use Llama or another model than you can control and host yourself, anything else is basically suicide.
You're still doing market research for OpenAI. Just because you aren't using their model doesn't mean they can't copy your UX. Prompts are not viable trade secrets after all.
No early stage start-up has revenues covering their expenses. But in fact you're right, it's not even “free”, it's “investor-subsidized” market research.
> You're still doing market research for OpenAI. Just because you aren't using their model doesn't mean they can't copy your UX. Prompts are not viable trade secrets after all.
Prompt aren't viable trade secret, but fine-tuning datasets and strategies, customer data[1], customer habits, user feedback, etc. are. And if you're using OpenAI, you're giving all that to them. Also, given their positioning, they cannot address any use-case that involve deploying your model inside your customer's infrastructure, so this kind of market research is completely irrelevant for them.
[1]: And don't get fooled by wordings saying that they don't train on customer data, they are still collecting much more info that what you'd like them to. For instance, even just knowing the context size that users like to work with in different scenario is a really interesting data for them, and you can be sure that they collect it and adapt to.
Code Interpreter was a pretty bad name (not exactly meaningful to anyone who hasn't studied computer science), but what's the new name? "advanced data analysis" isn't a name, it's a feature in a bullet point.
And it does seem "better" than standard 4 for normal tasks
What a terrible name! They should have asked ChatGPT for suggestions.
https://github.com/microsoft/azurechatgpt
Past discussion:
There are some great open-source projects in this space – not quite the same – many are focused on local LLMs like Llama2 or Code Llama which was released last week:
- https://github.com/jmorganca/ollama (download & run LLMs locally - I'm a maintainer)
- https://github.com/simonw/llm (access LLMs from the cli - cloud and local)
- https://github.com/oobabooga/text-generation-webui (a web ui w/ different backends)
- https://github.com/ggerganov/llama.cpp (fast local LLM runner)
- https://github.com/go-skynet/LocalAI (has an openai-compatible api)
Full Disclosure: This is my tool
Optimal business strategy. Makes it look like there's more competition, and changes the decision from "do we use ChatGPT" to "Which GPT vendor do we use?"
OpenAI is a tiny company, relative to Microsoft. They can’t afford to build a giant partner network. At best, they can offer a forum-supported set of products for the little guys and a richly supported enterprise suite. But the middle market will be Microsoft’s to own, as they always do.
Another question is does the latency even matter? Today, same employees ping their colleagues for answers and wait for hours till get a reply. GPT would be faster (and likely more accurate) in most of those cases.
I can actually see this saving a lot of time for employees (1-10% maybe?), so the price is most likely calculated on that and a few other factors. I think most big orgs will eat it like cake.
I’ve used it extensively to speed up the process of making presentations, drafting emails, naming things, rubber-ducking for coding, etc.
> Contact sales
Oops. Scary.
I'm missing the Teams plan: transparent pricing with a common admin console for our team. Yes, fast GPT-4, 32k context, templates, API credits... they're all very nice-to-haves, but just the common company console would be crucial for onboarding and scaling-up our team and needs without the big-bang "enter-pricey" stuff.
But hopefully it does give a little more motivation to all of the other great work going on with open models to keep trying to catch up.
I refer to the concept that output could be deemed free of copyright because they are not created by a human author, or that derivative works can be potential liabilities because they resemble works that were used for training data or whatnot (and we have no idea what was really used to train).
There was the recent court decision confirming:
https://www.cooley.com/news/insight/2023/2023-08-24-district...
Seems odd to start making AI systems a data-at-the-center-of-the-company technology when such basic issues exist.
Is this not a concern anymore?
There’s undeniably similar amounts of greed, although TK seems to genuinely enjoy being a bully versus sama is more of a futurist.
This decision seems specifically about whether the ai itself can hold the copyright as work for hire, not whether output generated by ML models can be copyrighted.
It's a pretty strange ruling at odds with precedent, and it has not been tested in court.
Traditionally all that's required for copyrightability is a "minimal creative spark", i.e. the barest evidence that some human creativity was involved in creating the work. There really hasn't traditionally been any lower bound on how "minimal" the "spark" need be -- flick a dot of paint at a canvas, snap a photo without looking at what you're photographing, it doesn't matter as long as a human initiated the work somehow.
However, the Copyright Office contends that AI-generated text and images do not contain a minimal creative spark:
https://www.copyright.gov/ai/ai_policy_guidance.pdf
This is obviously asinine. Typing in "a smiling dolphin" on Midjourney and getting an image of a smiling dolphin is clearly not a program "operat[ing] randomly or automatically without any creative input or intervention from a human author".
If our laws have meaning, it will be overruled in court.
Of course, judges are also susceptible to the marketing-driven idea that Artificial Intelligence is a separate being, a translucent stock photo robot with glowing blue wiring that thinks up ideas independently instead of software you must run with a creative input. So there's no guarantee sanity will prevail.
This is why music engravers can sell entire books of classical sheet music from *public domain* works. They become the owners of that specific expression. (Their arrangement, font choice, page layout, etc)
If the AI content is public domain, and the work it generates is incorporated into some other work, the entity doing the incorporation owns the work. It’s not permanently tainted or something as far as I know.
The only think that the ruling said is basically that the most low effort version of AI does not have copyright protection.
IE, if you just go into midjourney and type in "super cool anime girl!" and thats it, the results are not protected.
But there is so much more you can do. For example, you can generate an image, and then change it. The resulting images would be protected due to you adding the human input to it.
In practice, if you suspect something was written by an AI and are considering copying it, you would be safer to just ask an AI to write you one as well.
We'd be potentially very interested in an official internal-facing ChatGPT, with a caution that the economics of the consumption-based model have so far been advantageous to us, rather than a flat fee per user per month. I can say that based on current usage, we are not spending anywhere close to $20 per user per month across all of our staff.
[1] We used this: https://github.com/dotneet/smart-chatbot-ui
https://blogs.microsoft.com/blog/2023/07/18/furthering-our-a...
https://embracethered.com/blog/posts/2023/chatgpt-webpilot-d...
Seems like a no-go for companies if an attacker can steal stuff.
It seems like there's far more ways for extensive automation by decision making via ChatGPT to cause problems than solve problems. With my experience using it for programming, anything that is non-trivial basically requires you to do it all yourself because ChatGPT will hopelessly get it wrong.
By the way, I don't have to win a lawsuit to get some justice; I'll make discovery hurt.
So if OpenAI stole my IP and used it for training (which they probably did, illegally IMO), I guess you're taking that risk if you let your employees use their LLM's.
In the US the saying is "This case will come down to who has the last dime and it won't be you".
> By the way, I don't have to win a lawsuit to get some justice; I'll make discovery hurt.
Even for discovery a common tactic deployed for discovery is to "bury" the other party in discovery materials to the point where they will incur substantial costs just to parse the absurd mountains of materials they send you. Another "bleed you out" strategy.
Well, other than the millions of jobs at stake here. But I’m sure they can just learn to code or become an engineer
Which is a shame because an actual audit of the live training data of these systems could be possible, albeit imperfect. Setup an independent third party audit firm that gets daily access to a randomly chosen slice of the training data and check its source. Something along those lines would give some actual teeth to these statements about data privacy or data segmentation.
It's like slurping the very last capital a worker has out of its mind and soul. Most companies exist to make a profit, not to employ humans.
Paired with the pseudo-mocked-tech-bro self-peddling BS this announcement reads like dystopia to me. Not that technological progress is bad, but technology should empower users (for real, by giving them more control) not increase power imbalances.
Let's see how many people who cheered today will cheer just as happily in 2028. My bet: just a few.
Honestly, I can see it, but there are definitely SOME jobs at risk, and it will almost certainly reduce hiring in junior positions.
I am a manager in a dev team. I have a small team and too many plates spinning, and I’ve been crying out for more hires for years.
I moved to using AI a lot more. ChatGPT and Copilot for general dev stuff, and I’m experimenting with local llama-based models too. It’s not that Im getting these things to fill any one role, but to reduce the burden on the roles we have. Honestly, as things stand, I’m not crying out for more hires any more.
We already have enormous concentration of data in a few places and it's only getting worse. Centralization is efficiency, but the benefits of that get skimmed disproportionally, to the detriment of what allowed these systems to emerge in the first place: our society.
Technology should make life easier.
Automation is good.
A totalitarian state can make your life very comfortable with technology. Wanna trade for freedom?
Automation is the best, if the majority can benefit from it.
OpenAI loves to talk about a utopia where no one has to work and everyone is paid in Worldcoin (which of course Altman will make a handsome profit off of), but does anyone actually think that GPT-X is leading to this? Some of our most vulnerable members of society will soon find themselves without work, and no easy way to get new work. We don't need to wonder how we as a society will take care of them - all historic evidence points to us doing absolutely nothing.
I bet there will be a lot more effort to build truly open-source models. Also, I wonder why no foundation has yet got involved to pool resources and create large enough models.
It feels like AI needs its own "CNCF".
I'm very impressed by how Meta position itself as "AI for rest of us" position through llama and to an extent PyTorch, although I have no idea how they are going to capitalize that position besides hiring. (vs. Google having a cloud offering.)
Now you can pay real money for a chatbot to make stuff up about your company and its products.
While it is pretty incredible stuff, until or unless they have a veracity bit — some sort of "please don't lie" flag in it, I'd be wary of what it produces.
Will ChatGPT offer off-the-cuff pricing, discounts, rebates and refunds that are in line with your actual business model? Will it make written offers you will be legally-obligated to adhere to? Will it lie about your features and capabilities, or that of your competitors? Will it invent whole-cloth SLAs and KPIs you'll never be able to adhere to? Will it make up arrival and departure times? Medical conditions and prescribed treatments/therapies? Will it glibly give instructions that can get consumers/users who read these statements fired from their jobs, fined, imprisoned or killed?
There are places where you only want to give a generative AI so much room to color outside the lines. And after that, you are going to want traditional procedural logic to take over. Decision-trees from manually-maintained systems and even multi-decadal-honed expert systems.
There are some things where letting a computer make up a plausible response is ... okay-ish. And there are times where you can cause immense amounts of damage by letting things get outside of your control and human review.
This should come with a consumer warning label that would put your typical pharmaceutical safety disclosure to shame.
Now, with all that said, I am amazed at many of the results these LLM-based systems can generate. My main concern is that they are getting so good — or at least so plausible — that they can even fool domain experts unless they read really, really closely. And the experts won't be able to do that at scales of 10s or 100s of thousands of transactions per second.
Also I think you need to better understand how much hallucination has been driven down recently and its clear path going forward.
Like think of a technical B2B product's technical support engineer who now has a personal assistant trained on the company's full catalog of technical documentation, who can provide instant answers to customer questions that the rep can validate before passing along to the customer on the phone.
Or an overworked public defender who now has an instant paralegal to proof documents or search for and summarize relevant case law that the human lawyer then reviews.
And so on.
The statement you provided raises valid concerns about the capabilities and potential risks of advanced language models like ChatGPT Enterprise. Here's a response that addresses these concerns:
"Thank you for sharing your thoughts on ChatGPT Enterprise. It's important to recognize that advanced language models like ChatGPT do indeed have a remarkable capacity to generate human-like responses. However, they lack inherent understanding, critical thinking, and ethical judgment.
OpenAI acknowledges the concerns you've highlighted about accuracy, veracity, and potential consequences of generated content. It's crucial to understand that ChatGPT does not have an inherent "veracity bit" or a built-in mechanism to discern truth from falsehood. As such, it's important for users to exercise caution and discretion when relying on the output generated by the model.
In situations where accuracy and compliance are paramount, it's advisable to rely on established procedural logic, expert systems, and human review. ChatGPT is not a substitute for human oversight, particularly in sensitive areas such as legal, medical, or business transactions.
OpenAI is actively working to improve the control and guidance mechanisms for ChatGPT, and they acknowledge the importance of striking the right balance between automation and human review. As you pointed out, there are instances where generative AI is appropriate and others where traditional procedural logic and human intervention are necessary.
While ChatGPT can be a powerful tool for generating content and ideas, it is incumbent upon users to apply their expertise and judgment to evaluate the output, particularly in contexts where accuracy and adherence to policies are critical.
Your concerns align with ongoing discussions in the AI community about the ethical use and potential limitations of these technologies. As AI systems continue to advance, responsible development, transparency, and user education will remain important factors in their deployment."
It's important to remember that while AI models like ChatGPT have the potential to assist and enhance various tasks, they do not replace the need for human expertise, ethical considerations, and careful oversight in complex and critical scenarios.
Even if no GPU is simultaneously being used for enterprise and normal, they still have a finite amount of compute.
How can we be sure of this? Just take their word for it?
If you notice that some of your confidential info made it into next generations of the model, you'll be able to sue them for big $$$ for breach of contract. That's a pretty good incentive for them not to play stupid games with that.
Yeah... I have no doubt that people at my Fortune 100 company tried it out with their corporate email domains. We have about 80,000 employees, so it seems nearly impossible that somebody wouldn't have tried it.
But, since then the policy has come down that nobody is allowed to use any sort of AI/LLM without written authorization from both Legal and someone at the C-suite level. The main concern is we might inadvertently use someone else's IP without authorization.
I have no idea how many other Fortune companies have implemented similar policies, but it does call the 80% number into question for me.
Surprising it didn't make the initial launch announcement though.
Why can't I, as an individual, have the same features of an Enterprise plan?
What is the logic behind this practice other than profit maximization?
I'm willing to pay more to have unlimited high-speed GTP4 and Longer inputs with 32k token context.
EDIT: since I'm getting a lot of replies. Genuine question: how should I move to get a reasonable price as an individual for unlimited high-speed gpt4 and longer token context?
That’s a real big “other than”…
How much more? That’s the question that “talk to us” enterprise pricing is trying to answer.
I don't know, but I can't imagine any other logic.
Maybe posting the price they'd like to charge would scare away almost all interested parties.
Maybe the price they charge you depends more on how much money they think you have than it does on a market's "decision" on what the product is worth.
Why would it be something else than profit maximization? It's a for-profit company, with stakeholders who want to maximize the possible profits coming from it, seems simple enough to grok, especially for users on S̵t̵a̵r̵t̵u̵p̵ ̵N̵e̵w̵s̵ Hacker News.
https://blogs.microsoft.com/blog/2023/07/18/furthering-our-a...
Are you aware what the entire point of a business is?
# OpenAI Offerings
- ChatGPT Free - trains on your data unless you Opt Out
- ChatGPT Plus - trains on your data unless you Opt Out
- ChatGPT Enterprise - does _not_ train on your data
- OpenAI API - does _not_ train on your data
# Microsoft Offerings
- GitHub Copilot - trains on your data
- GitHub Copilot for Business - does _not_ train on your data
- Bing Chat - trains on your data
- Bing Chat Enterprise - does _not_ train on your data
- Microsoft 365 Copilot - does _not_ train on your data
- Azure OpenAI Service - does _not_ train on your data
Opt-out link: https://help.openai.com/en/articles/7039943-data-usage-for-c...
And if you opt out, they delete the chats from your history so you can't reference them later for your own use. Slick!
doesn't say they're not selling it to someone else who might...
aka the nerfed version. high speed means the weights were relaxed leading to faster output but worse reasoning and memory.
me: “I would like to close my account”
chatgpt: “I’m sorry, did you mean open or close an account”
me: “ close account”
Chatgpt: “okay what type of account would you like to open”
Me: “fuck you”
Chatgpt: “I’m sorry I do not recognize that account type. Please repeat”
me: “I would like to close my account”
Chatgpt: “okay i can close out your account,please verify identity”
me: <identity phrase>
Chatgpt: I’m sorry that’s incorrect. Your account has been locked indefinitely until it can be reviewed manually. Please wait for 5-10 business days
I'm sorry but the financial analyst using chatGPT to write their excel formulas for them, and explicitly calling out that it is generating a formula that the analyst can't figure out on their own ("tricky") is an incredibly alarming thing to call out as a use case for chatGPT. I can't think of a lower reward, higher risk task to throw chatGPT at than financial analysis / reporting. Subtle differences in how things are reported in the financial world can really matter
This sound like a huge waste of money for something that should just be completely on-device or self-hosted if you don't trust cloud-based AI models like ChatGPT Enterprise and want it all private and low cost.
But either way, Meta seems to be already at the finish line in this race and there is more to AI than the LLM hype.
I can imagine this argument being made repeatedly over the past several decades whenever anyone makes a decision to use any paid cloud service. There is a value in self-hosting FOSS services and managing it in house and there is a value in letting someone else manage it for you. Ultimately it depends on the business use case and how much effort / risk you are willing to handle.
From the very beginning everyone knew data privacy & security would be one of the main issues for corporations.
We're going after some of these use cases:
Want a daily email with all the latest news from your custom data source (or Google) for a topic? How about parsing custom data and scores from your datasets using prompts with all the complicated bits handled for you, then downloading as a simple CSV? Or even simply bulk generating content, such as generating Press Releases from your documents?
All easy with FlowChai :)
I think there's room for many different options in this space, whether that be Personal, Small Business or Enterprise.
Here's an example of automatically scraped arXiv papers on GPT4, turned into a report (with sources) generated by GPT4: https://flowch.ai/shared/6107d220-4e19-4bdc-a566-e84e8a60565...
I quick scrolled through your webpage and had no idea what it was. Extremely text heavy, and generic images that didn't communicate anything. I wanted to know what the product LOOKED like, especially as you're describing the difference between it and the chat interface of OpenAI.
I think you updated your comment (or I missed it) with the link to a "report" - it looked just like the output of one of the text bubbles except it had some (source) links (which I think Bing does as well)? It didn't seem all that different to me.
You didn't offer me any way to delete my account and remove the email address I saved in your system. I hope you don't start sending me emails, after not giving me an ability to delete the account