Don't Pay For ChatGPT Pro? Unlimited Access To OpenAI o1, o1-mini, GPT-4o, and Advanced Voice
Вставка
- Опубліковано 13 гру 2024
- ChatGPT Pro is a subscription plan priced at $200 per month, offering unlimited access to OpenAI’s advanced models, including o1, o1-mini, GPT-4o, and Advanced Voice. The plan features o1 pro mode, which utilizes additional computational power to deliver more accurate and comprehensive responses, particularly in areas such as data science, programming, and case law analysis.
SUBSCRIBE for more! 👉 bit.ly/3zlUmiS 👈
Introducing ChatGPT Pro
Broadening usage of frontier AI.
openai.com/ind...
👨💻 Ask Me Anything / Exclusive Content (119+ Lessons) ☕
www.skool.com/...
💬 Chat with this UA-cam video
bumpups.com/
-------------------------------------------------
➤ Follow @webcafeai
• 🤠 2nd Channel (digital nomad): / @corbinwander
• ✖️ X: x.com/webcafeai
• 🥾 Instagram: / webcafeai
• 🎧 Bräunlich: / braunlich
-------------------------------------------------
Key Takeaways:
✩ Subscription Details: ChatGPT Pro costs $200 monthly and provides unlimited access to OpenAI’s top-tier models, including o1, o1-mini, GPT-4o, and Advanced Voice.
✩ The Pro plan includes o1 pro mode, which offers increased computational power for more reliable and detailed responses in complex fields like data science and programming.
▼ Extra Links of Interest:
automate everything. 👇
linktr.ee/webcafe
My Setup To Record Content (amazon storefront) 📷
amzn.to/4d8Qkg2
Become an Early Adopter 🍻
/ @corbin_brown
My name is Corbin, creator of bumpups.com and investor behind Webcafe AI.
I build things for fun 🤠
If there are differences, they will come out when you give hard, expert level problems.
Comparing the solutions to problems that 4o or o1-preview could already solve to a satisfactory degree doesn't make sense.
Afgreed I thought the same. He didn't really test it to it's capacity and I'd say not even close.
Maybe feed it files and data and provide a set of proper instructions, then manually validate the output. Maybe you'd get a more qualitative test that is relevant..smh
but I appreaciate the video nonetheless makes me want to give it a shot
was coming to type that. As a developer, I like to upload heavy and complicated documentation sometimes and o1 baby omg so good.
The title dripped me into thinking how to get advanced voice mode for free and that’s not what the video was about
Yeah I unsubbed due to the clickbait title.
"It's not clickbait, @legendfpv. I have the Plus, so trust me, I understand the title better than you."
The plans keep saying "limited access to the o1 model" but they don't really say HOW limited. Like, are they reducing how many tokens I can use? By how much?
if ur on the plus plan 50 messages per week pro is unlimited and this is the app/web platform idk about apis or other
you get 50 o1-mini a day, and 50 o1 a week, pro gets unlimited access to almost everything but they
get limited to o1 pro mode, though they are a tad more vague as to what the limit of o1 pro mode is.
Been using pro. You have to prompt it entirely different than even regular o1 to get the best performance from it.
Do you have any examples? I've been contemplating to whether to upgrade or not
can the pro mode read long document like 4o?
Bro just a price of advice. Make the title shorter delete everything and make it “Is ChatGPT Pro 200$ worth it?” It looks way better like this.
I tested it for complex algorithm problems, and you can definitely see the difference, don’t worry about that
I'd love to watch video or quick comments on examples and visible performance differences 😊
Thanks for sharing such valuable information! I need some advice: I have a SafePal wallet with USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). Could you explain how to move them to Binance?
The 128k context window is a big thing too
I mean, the prompts you gave the model aren't good examples of this model's capabilities. This prompt will get correct answers 8 times out of 10 by the 4o model, but the answer is accurate. Maybe for someone doing state-of-the-art research, using O1 Pro might make sense, but if you are not facing limit problems, 4o and, in rare cases, O1 are more than enough for most people and they should not pay the extra 180.
the main reason is not how good the pro mode is. It's the unlimited usage of reasoning models.
If that's the case, very questionable if it's worth the money
These prompt examples weren’t complex enough to demonstrate the difference in capabilities between O1 and O1 Pro, hence the very similar results.
You’d have to build into a topic or have a much longer prompt that provides greater detail, probably on knowledge that it doesn’t already have to let it think and flex it’s reasoning capabilities
This dude is about to pitch his how to do algo trading bot
Better to give it proper coding problems, taking existing code and adding features, for example. Sonnet 3.5 is the only game in town right now. o1 doesn't cut it in coding. Pro might.
Thank you for your awesome comparison, But I believe it's not worth it.
You should read a few books on systems thinking and theory of constraints, learn the concept of building the logical trees, and then you will find out why this new pro model costs 10 times more than the plus.
What you show in this video makes no sense, because research is always done in multiple iterations. AI just helps you speed up the process, but you should do the "thinking" and "prompting" part.
Expecting to get genius output from a one sentence prompt is unrealistic...
thanks man!
Why ask silly simple questions when the o1 pro is built for more complexity? 4o can't do multi threaded reasoning, evaluating and combining the answers into a cohesive reply
First I have to earn $200 to pay it
i really don’t think o1 is all that great
You have to prompt it differently than classical models since it uses an entirely different paradigm, the more specific and precise you are the worse it tends to perform, meaning you should try to give it a clear picture of what you want but avoid trying to guide it in a step by step fashion.
@@JaceRivera-t2w Where is this been proven? Do you want to balance half-and-half? You don’t want to define it that it goes too narrow but I would say you also don’t want to keep it unconstrained and not get what you want.
@@augmentoslisten o1 uses test time compute and consensus based sampling. This means that when you provide a prompt o1 quite literally breaks up your prompt among multiple “instances” and then it uses unfiltered chains of thought that are allowed to be as creative as possible in solving the problem then each “instances” votes on the solution. This improves the reliability of the response.
Think about it like this in real life if your boss, manager, teacher, investors, etc are to rigid on how you perform a task then you run into an issue where you end up performing far worse than you would otherwise since your creative problem solving would be hindered.
You can also think of it like this as well
“Classical models are impulsive therefore you have to spell out everything for them, whereas o1 is creative and introspective thus you should provide it with clearly defined goals but be very vague as to how it should achieve those goals and only mention constraints that would effect the desire outcome. So if you are solving a physics problem and you want to make sure potential solutions are bound to non-zero real numbers then state that.”
If you prompt it like that you will be absolutely amazed
and it's the best one they got. 4o is incredibly bad.