GPT-4 VS. Gemini Ultra (The Ultimate Head to Head Comparison)
Vložit
- čas přidán 28. 05. 2024
- GPT-4 has been the dominant AI chatbot since it came out and no one has been able to create something that can outperform it just yet. Google Ultra / Google Advanced, the paid subscription from Google is the first AI model claiming to outperform GPT-4 in some benchmark testing.
I’m going to turn both models through a set of 10 prompts across the most commonly used tasks and I’ll assign a score of 1-5 to each based on the results. We’ll test writing, research, creativity, coding, reasoning and several more categories.
Usability test
Before I run the two prompts, let’s talk about a few limitation they both currently have.
ChatGPT has a limitation in how many answers it can provide. Gemini currently has no limitation.
Price - As far as price, they are both $20 dollars a month, but with Gemini, you also get a few other benefits like 2 TB of storage.
Privacy - Both chatbots use conversation for training, but ChatGPT has a teams plan that will keep the conversations private.
Speed - When running both through a speed text, Gemini outperforms GPT-4.
Context Window - GPT-4 has a context window of 32k and with the turbo model, 128k.
1 - Text summarization
Prompt: Provide two summaries of this article. The first summary should be 2-3 sentences long. The second summary should be 5-6 sentences long and include more detail.
2 - Writing Text (Tone, style and accuracy
Prompt 1: Concise Product Description
Topic: Imagine you're launching a new software tool that helps businesses track social media analytics.
Prompt: Write a short, punchy product description (approximately 50 words) suitable for a website or marketing material. Emphasize the key benefit for businesses.
Prompt 2: Email Tone Shift
Topic: A colleague has sent you a frustrated email about a delayed project.
Prompt: Draft a reply that acknowledges their frustration, offers a solution-focused approach, and maintains a professional but reassuring tone.
3 - Multimodal Understanding
Prompt: Describe what's happening in this image and explain the likely cause of the person's frustration. (Upload image)
Prompt: Turn this image into html and css code I can copy and paste. (Upload screenshot)
4 - Image generation
Prompt: A photorealistic image of a cat wearing sunglasses and riding a skateboard.
Prompt: Design a powerful logo for a construction company called 'Apex Builders'. The primary focus should be bold typography, possibly incorporating a subtle construction-related graphic
5 - Research (completeness and accuracy)
Prompt: Research the concept of quantum entanglement. Provide a summary suitable for a high school student, and include links to two credible sources that explain it further.
Prompt: How could artificial intelligence (AI) potentially disrupt the accounting industry? Identify specific use cases, potential benefits, and challenges. Provide links to relevant articles or reports.
6 - Creativity
Prompt: Choose an everyday object (like a coffee mug, a paperclip, or a shoelace). Write a short monologue from the object's point of view, giving it a personality and a unique outlook on the world.
Prompt: Write the opening scene of a story that blends film noir with high fantasy. Think gritty detective meets magical creatures and quests.
7 - code generation
Write Python code for a game of snake that I can run on my computer.
Then tell me the step by step guide on how to do it, assuming I know nothing about programming.
8 - Complex reasoning
What has a voice but cannot speak? It has a bed but never sleeps, it has a mouth but never eats, and it runs but has no feet.
At a party, each guest shakes hands with every other guest exactly once. If there were a total of 66 handshakes, how many guests were at the party?
9 - capabilities (Plugins and Extensions)
Gemini currently has 5 extensions but they are actually pretty powerful, special CZcams and Google workspace.
However, GPT-4 has the entire GPT store with thousands of custom GPTs.
10 - Content creation
Prompt:
Task: Extract the core lesson or actionable tip from the CZcams script. Condense it into a concise tweet or LinkedIn post suitable for quick consumption.
Example: "[Video Title] boiled down to one key idea: to boost focus, try shorter work sprints, not longer ones. #productivity"
Master ChatGPT, Midjourney, and top 50 AI tools with Our New AI Education Platform.
Start a free trial today: bit.ly/skillleap
Introduction (0:00 - 1:02)
Briefly introduces GPT-4 and Gemini Ultra
Mentions the purpose of the video: to compare the two AI chatbots
Limitations (1:02 - 3:18)
Discusses limitations of GPT-4: cost, privacy, context window
Mentions that Gemini Ultra does not have limitations on usage
Summarization (3:18 - 5:22)
Tests how well both AI chatbots can summarize text
Provides prompts and compares the quality of the summaries
Concludes that both GPT-4 and Gemini Ultra do an excellent job summarizing text (scores them both 5 stars)
Writing (5:22 - 7:42)
Tests how well both AI chatbots can write different kinds of copy (e.g., product descriptions, emails)
Provides prompts and compares the quality of the writing
Concludes that Gemini Ultra is better at providing different creative options while GPT-4 is better with following instructions and staying within the word limit (gives GPT-4 4 stars and Gemini Ultra 3 stars)
Code (7:42 - 9:22)
Tests how well both AI chatbots can generate code from an image
Provides a prompt and compares the results
Concludes that GPT-4 is better at understanding and generating code (gives GPT-4 4 stars and Gemini Ultra 2 stars)
Image generation (9:22 - 11:22)
Tests how well both AI chatbots can generate images from a prompt
Provides prompts and compares the quality of the images
Concludes that neither GPT-4 nor Gemini Ultra is perfect at generating photorealistic images, but GPT-4's images are better formatted (gives GPT-4 0 stars and Gemini Ultra 0 stars)
Research (11:22 - 13:22)
Tests how well both AI chatbots can find information and answer questions based on a prompt
Provides prompts and compares the quality of the answers
Concludes that GPT-4 is better at finding relevant information and providing links, but both models sometimes struggle (gives GPT-4 2 stars and Gemini Ultra 1 star)
Creativity (13:22 - 15:22)
Tests how well both AI chatbots can be creative
Provides prompts and compares the quality of the creative text formats
Concludes that both models are similar in creativity (gives GPT-4 3 stars and Gemini Ultra 3 stars)
Content repurposing (15:22 - 17:22)
Tests how well both AI chatbots can help repurpose content for different platforms
Provides prompts and compares the quality of the repurposed content
Concludes that both models need work on repurposing content, but GPT-4 might be slightly better with the help of custom GPTs (gives GPT-4 3 stars and Gemini Ultra 3 stars)
Conclusion (17:22 - 18:57)
Summarizes the findings of the comparison
Concludes that GPT-4 is still the better AI chatbot overall, but Gemini Ultra is a promising new model with a lot of potential
Outro (18:57 - 19:10)
Calls to action: subscribe to the channel, enroll in the e-learning platform
00:01 Google Gemini Ultra challenges the dominance of GPT-4 in AI chatbot performance.
02:04 Privacy and context window comparison between GPT-4 and Gemini Ultra
05:58 Both GPT-4 and Gemini Ultra receive five stars for summarizing text.
07:55 Gemini excels in writing email copy with great length and formatting
11:56 Google Gemini is advancing multimodality but needs improvement
13:50 GPT-4 and Gemini Ultra are not effective for image creation
17:47 AI in accounting use case and challenges
19:33 GPT-4 excels in creativity with comprehensive and well-formatted output.
22:58 GPT-4 is a powerful coding assistant for non-developers.
24:55 Google Gemini and Gemini Ultra's capabilities and performance in reasoning and complex reasoning.
28:46 GPT-4 offers a variety of custom GPTs and potential for future growth.
30:33 GPT-4 VS. Gemini Ultra for content creation on social media
34:00 Gemini Ultra has some beta features but has potential to surpass GPT-4
I confess that when I showed both Chatgpt4 and Gemini Advanced a diagram of an electrical circuit and asked them to calculate the total capacitance. . Chatgpt4 got it right and Gemini Advanced got it Wrong. Gemini Advanced is good at analysing the grammatical structure of Spanish sentences. I admit Chatgpt4 crashed a few hours ago when trying to do calculations about another circuit. .this time Gemini Advanced DID answer the questions about the circuit correctly but I had to describe the circuit in great detail.
Gemini Advanced currently uses Google Lens to decipher images. It has not got it's built-in multimodal capability.
Great video. For me, I just overwhelmingly prefer the writing style of Gemini Ultra. Since most of my use cases revolve around writing, I've basically given up using ChatGPT4. One area where I disagree with the video is that I overwhelmingly prefer Gemini advanced for creative writing as well. GPT just has a tendency to always lapse into a style that reeks of a college student trying to show off how intelligent it is. It has a pompous over the top formalistic essay vibe to everything it writes. Most of the time Gemini Advanced literally sounds human to me when it writes something. Most of its output is just imbued with sort of a warmth and Humanity where GPT just seems to be very cold and formal.
I've also grown addicted to the options of being able to choose if you want the output longer or shorter as well as having three drafts of everything and on top of that having unlimited prompts and the access to CZcams.
I never realized how constraining being limited to 40 problems every 3 hours was until I had got rid of it. Anyway, there's no right and wrong, it's obviously subjective, but for me for the actual output and the functionality and even the user interface I've come to overwhelmingly prefer Gemini Advanced at this point. Which is shocking to me since I've been absolutely addicted to chat gpt4 and brag about it to everyone for the past year.
Edited with Ai
True, I've swapped over my GPT-4 subscription to Gemini Advanced and I'm giving it a try for 2 months. So far, everything excels much better for me and I'm really loving the integration with google apps since I do a lot of work in google docs and sheets. Only issue at the moment is the AI art stuff and the lack of custom chat AI like GPT-4 has, although I have no doubt they're gonna make this better with every update and once gemini 1.5 releases
True, GPT-4 is often terrible for writing just with how flowery it is if not controlled, like a 9th grader trying to impress the teacher (and fairly obvious AI writing in general). Also like Gemini's formatting and readability. Think it has a good chance to improve faster. Using both for now for different purposes.
That's because you have to tell it what you want personality
The comparison is truly helpful and very
detailed based on scenarios. Thanks
Thanks for doing detail comparison
WHEW! That was quite a detailed comparison. Thank you for this one.
Loved the new channel theme
Because you are one of the few folks that create professional comparisons: Could you create one comparing ChatGPT Plus with Copilot Pro? Thanks.
As a developer that used gpt4 ... The version before the November update was insane... Like most code you asked it got right in first try ... But now it's not perfect every time and you need to either fix few issues , or it doesn't get the question /prompt perfectly
oh good to know. thanks.
As a programmer / engineer, I have Gemini ultra because it’s cheaper, would you say that gpt4’s better capability is worth the higher price?
@@Dtthew I am not sure how good Gemini advanced is .. I use the normal version and gpt4 is still better for non programming tasks ...also it has a lot more bias than gpt4 due to restrictions put on it from Google... Eg. It sucks at even simple maths ... Either that or it forgets maths when it comes to controversial topics and can't even compute extrapolating a basic percentage value ...
Gemini Advanced is fabulous at coding, feels like they optimized it for it. My new favorite, it gives concise and bullet point explanations too which makes it easy to understand. It also keeps getting better and will get a python interpreter built in soon. Also 2TB of storage! Can’t wait for them to integrate google drive as location for code outputs.
Yo, absolutely!! It’s insane, it used to be amazing and now it just seems dumb, what do you use now for programming help?
I know you said it but Gemini was JUST released. IMO it looks promising as hell if you give it enough time to catch up and maybe even surpass the competition. Great video!
Great Comparation👊
I liked the video. Great content. The moving neon frame is distancing.
ok thanks. i've avoid it
@@SkillLeapAI :)
Great video. Well done sir!
I love yours videos. From my point of view, you are the best CZcamsr in the domain of AI. Thanks
Great video. I used GPT3.5 for 9 months, GPT4.0 for 3 months, Gemini advanced for 1 month. I have almost the exact comments you have in the video. GPT is much better in most perspectives. Gemini doesnt know how to debug and give up immediately when there is a bug in its code. Gemini says it cannot handle pictures with people inside. The pictures generated by Gemini can only just for fun and cannot be used for any meaningful usage. There are so many logical 'bugs' inside the picture. I have a picture a person standing outside the cage, I can still see some iron mesh on his face. but not on the body. There is only 1 area Gemini wins, as mentioned by you, GPT doesn't provide the links for reference; or the links do go anywhere.
Great video! What AI would you highly recommend right now for marketing purposes? Content creation, strategy, planning, copywriting for posts, creativity, etc. 🙏🏻
You mentioned you don't use either of thesevand are using Perplexity AI. Now I'm going to be searching for that to see if it's something I would rather have. I signed up for Chat GPT, the version that hasn't been updated since January 2022, so was looking for a more up to date version and came across your video. Glad that I did. Following you to see more soon. Cathy
I use perplexity just for research. I use these way more for other tasks like writing and analyzing documents. I'll make a video about why I use different apps for different use cases soon.
Can't wait to see which one wins.
I think it would be useful for other people if I gave information about generating images in addition to the video. I needed to create a logo for my own company and I tried to create it with Gemini Advanced. At first it refused to render as mentioned in the video and offered useful urls, but when I tried it a little harder and changed the prompt, it managed to create a logo that looked really nice and the text on it looked correct.
And in addition, thanks to the new update released to Gemini, it helps you understand and run Python codes much better.
YOO, why not demonstrate the full avengers line up with specialized GPT's of Chatgpt VS Gemini. and does gemini have any special capabilities?
Privacy you can opt out of training in gemini already, it’s in the app history thing, if you toggle it off for saving then it won’t be used for training
Oh perfect. Couldn’t find it before
Bro in every problem solving field that i've tested Gemini Ultra vs GPT 4 , GPT 4 just abolishes it in every field math , physics , chemistry , puzzles , computer coding like in every possible field .
Please also include the comparison of Microsoft copilot and Google Android Assistant with Chatgpt4 & Gemini AI ultra.
In my experience, ChatGPT excels at code generation, while Gemini shines in document tasks. While I found some outdated links in Gemini's output, it seems more an issue of link drift than generation. As a software engineer, I leverage both tools for their strengths!
Great information, thanks. There's a lot of AI's tool popping up and little time to digest every single one. In your opinion what is the best FREE AI tool to get graphics without going into a lengthy learning curve? Thank you very much.
I would say dalle inside of copilot. It’s basically a free way to access dalle 3. Adobe express is also good for more complete graphics
GPT was incredible when released, now it feels dumb af and contantly rejects your request or give you half answers
Thanks, does mid journey creates better logo and add text properly?
Midjourney doesn’t do as well with text. It’s pretty good with logos though
10:42 It's not that Gemini can't tell you if there is an image of a person it's just that Gemini won't even see the image in the first place. The Underlying software running on "Top" of Gemini will automatically remove the image from your prompt if a person is detected and give you an apologetic response.
People think you get to play with the model when you pay for it. But what you actually play with is a PrePrompted Chat-Augmented Customized-Version of the Original Model.
TL: DR
Gemini can explain to you images of People, but Google won't let it see the image.
That was great. How about copilot pro.
Do you recommend chat gpt 'plus' or gemini advanced for the long run?
hard to say on the long run. They change and evolve so fast. GPT-4 is still my go to at the moment
How are you today my friend ?I want to make small cartoon videos but I want to know something please since the video shot is a group of photos , frames Is there a way in Midjourney to give it a photo reference and the action I want to be made and Midjourney makes let's say 30 consistent photos or frames to make the shot, I take the 30 photos and put it in any video editing program and make the scene? etc. etc. thank you.
Not exactly. But you can create the first frame in midjourney and then use runway to turn it into a video clip. That’s how we make our short films. I made a video on the process last month on this channel
yeah, I've seen it but unfortunately runway is only 4 seconds long, and this is too short and not good with legs and moving things!@@SkillLeapAI
i've just started watching and i will keep watching, so maybe you correct this later. The context window is nothing to do with how much you can put in a prompt. It may seem a small distinction but is far more important than you might realize. The context window is how much the serverside app can put into an LLM prompt. A small error people make but one that creates a large error in perception when looking at capabilities.
Good point. I think I simplify it a little too much
In Chatgpt u can make ur own GPT’s. I use it for uploading manuals, so I can ask it about them. Can Gemini do something similar?
No they don’t have that yet.
Thanks for this informative video. It's definitely worth of 66 high fives!😂
I am very confused sir. What Ai chatbot should i buy for everything? chatgpt, gemini, copilot, claude 3, or any other
gpt is still far better it was developed far longer and i feel like gemini is rushed, it my case sometimes it's having a hard time debugging a problem
Gimini Advanced can not generate images. I tried different prompts versions . Last time, it suggested that I ask DALL-E or Midjourney. 😂 How do you do that? 😅
They paused it at the moment after I made this
As a programmer I agree with the coding for assessment. I would rank GPT a 3 based on my tests. It is great with python and html coding. It does fail on more advanced coding tasks. The snake game is impersive when I did my testing it was the only llm that created a successful python snake game (on the first try) where other models failed even with extra prompting. It does fall short with advanced coding concepts. I tried to get a step by step guide for creating a snake game in Unity. It did do a good job and it did create a somewhat functional game, it failed in coming up on it's own that I should use some of the built in components instead. I had an issue with the snake growing properly after getting stuck in a loop with text, I took screen shots of what it was doing and did a rough drawing of what I wanted it to do. It did describe the problem perfectly along with exactly why it was failing, but failed to come up with a proper solution it just got stuck in a loop. I did get it to get to the correct solution when I gave it a specific solution using corutines at the end of the test. It did an awesome job with the rest of the game and it was great at giving me a step by step guide. It was also able to generate some good images that could be used for the game, it did have an issue creating the snake but after several tries it did give me something that was usable. The first image it creates for each concept it gives you one image but if you tell it that you don't like that image to try again and it will give you 4 images, it did a good job at refining that image. I told it image 2 was close to what I wanted but it wasn't a top down image, so it produced 4 more images and I tell it this is I really like image 3 but can you remove xyz or change the shape of the head to be like image 4. And it was able to do it. Using mid journey or stable diffusion gives better results, but I did find it easier to refine an image to be exactly what I imagined with chat gpt. Chat gpt also did an awesome job for creating python scripts that you can use in blender to help create basic shapes or do some advanced refining of models in blender For coding Chat Gpt does outperform most llms out there but there is a lot of room for improvement. Code pilot does a better job then Chat GPT, even though it uses the same model, it just uses a slightly more refined version of that model, I am guessing that might have to do with the licensing and code pilot actually being able to use more resources for examples of code where Chat gpt and Google are more generic everything models.
GPT stands for Generative Pre-training Transformer. In essence, GPT is a kind of artificial intelligence (AI).
Really insightful comparison between GPT-4 and Google Ultra. For entrepreneurs leveraging AI for lead gen, remember that the speed and context window of your AI tool can significantly impact your outreach personalization and follow-up strategies, especially when dealing with large datasets. Advanced users can integrate AI to not only initiate contact but also to manage and predict lead scoring based on interaction data.
Um sorry but did you change your mic cuz your voice literally sounds more robotic
No same mic I’ve had for a year. Let me check my settings
Wonder if they'll wrestle to decide the winner?
porque chatgpt 4 plus entrega imagens em webp e nao em png?
I’m not sure why they changed it recently
Did you ever do a video on huggingchat?I feel something special about it@@SkillLeapAI
You gotta respect the smack talk.
I just signed up for Gemini Advanced. How is Gemini Ultra different from that?
It's the same. Gemini Ultra is the model that powers Gemini Advanced. The regular Gemini is powered by Gemini 1.5 which is not as good as Ultra 1.0. They really should name Gemini Advanced Gemini Ultra. It's confusing.
@@SkillLeapAI you don't know what you are talking about. Gemini 1.5 pro is not a released model. Gemino 1 pro powers the free version of bard. Gemini 1 ultra powers Gemini advanced but its multimodal capabilities are not enabled.
Would you like to be a judge?
1. CodeCraft Duel: Super Agent Showdown
2. Pixel Pioneers: Super Agent AI Clash
3. Digital Duel: LLM Super Agents Battle
4. Byte Battle Royale: Dueling LLM Agents
5. AI Code Clash: Super Agent Showdown
6. CodeCraft Combat: Super Agent Edition
7. Digital Duel: Super Agent AI Battle
8. Pixel Pioneers: LLM Super Agent Showdown
9. Byte Battle Royale: Super Agent AI Combat
10. AI Code Clash: Dueling Super Agents Edition
Why and how can you even compare the two? Gemini refuses to answer most questions
I find Gemini much more organized, but GPT-4 is more precise and accurate.
📝 Summary of Key Points:
📌 The comparison between GPT-4 and Google Gemini Ultra in various categories like text summarization, writing, research, creativity, coding, reasoning, and more.
🧐 Limitations and usability issues of GPT-4, including subscription plans, privacy concerns, and context window capabilities.
💡 Additional Insights and Observations:
💬 Quotable Moments: Gemini Ultra showed promise in some areas like image generation and CZcams integration, while GPT-4 excelled in creativity and coding tasks.
📊 Data and Statistics: Gemini Ultra's extensions and CZcams integration were highlighted as potential game-changers in the AI chatbot space.
🌐 References and Sources: The comparison emphasized the strengths and weaknesses of both GPT-4 and Gemini Ultra in different functionalities.
📣 Concluding Remarks:
The video provided a detailed comparison between GPT-4 and Google Gemini Ultra, showcasing their strengths and weaknesses across various categories. While GPT-4 excelled in creativity and coding tasks, Gemini Ultra showed potential with its extensions and CZcams integration. Both AI chatbots have room for improvement, but the competition seems to be heating up in the AI space.
Generated using TalkBud
If it’s just pure gpt-4 vs gemini why not use Copilot Pro? No limitations, and even though dall-e says 100 uses a day it’s actually a lot more, that’s just how many boosts you have, plus like gemini advanced it searches way more.
Yea I should make that video as well
Co pilot sucks
copilot is rly better than gpt4 and gemini advanced@@MustardGamings
I agree gpt4 is an absolute monster for coding, it’s helped me bridge the gap in my skills, I feel so powerful with what I can do coding wise with its assistance
Is it just me or as of recent, Chat GPT 4 doesn't get web access???
It should have browsing with Bing
The comparison should have been between OpenAI best ( teams plan ) vs Gemini best ( advanced )
I guess technically I was using the teams plan. I don’t have anything else. But teams and regular use GPT-4. Slightly different models. Only other thing with teams is privacy and higher limitations on chats
You forgot to share the most important point. The data from chatGpT goes back until 2021. Google Gemini on the other hand, data is from yesterday.
That’s not correct. That’s the much older ChatGPT. They’ve updated to Dec 2023 and GPT 4 has web access with Bing too
Correction: the GPT 4 Turbo 128k model is default to plus user … the more resource hungry GPT 4 model but in the 32k version is used in teams. So, teams users are getting the better model… probably the main reason why we as teams users haven’t experienced any strangeness in todays GPT Turbo hiccup. 😅
Plus users don't have 128k lol. Both have 32k version
@@MustardGamings Are you certain… I'd check again. To my knowledge, GPT 4 Turbo has 128k context (read) with 4k completion. Then there is the more capable GPT 4 32k variant we have in teams. I can't paste links here.
@@MustardGamings Technically both Plus users and Teams have 128k context. However, the ChatGPT website limits the maximum tokens to 32k, to maintain speed. but the underlying model accepts 128K
Hi i'm a french frog and studying python in 1st year univ i see free version gemini pseudocode sucks it's a melange between python and pseudocode it's not for real pseudocode with university or pro grade but his python synthaxe is more elegant.
Gemini is easier to navigate and use code implementation they have the chrome experience in fluidity of navigation it's logical but chat gpt color police and black background is a lot better to read for analyse code.
In math and discretes maths chatgpt make a little less errors but it's very close and free version of gpt havent pictures ia analysis and ocr like gemini both are complementary i love them they help me a lot of algorithms of sort who are a pain to ass to learn and understand for beginners
Free Gemini does not create imges for me ?
It doesn’t? It should. It’s not limited to paid version
Not available in Europe currently.
Even though on paper ChatGPT4 and Gemini Advance have the same 32K context window, I can copy-paste much more text as a part of a prompt into ChatGPT4. I wonder why?
Yes. I pasted 3X the content into GPT4 as Gemini Advanced. Otherwise, I like the writing style of Gemini better than then often poetic and prissy verbose writing of GPT4 :)
Very good video, thanks for sharing. Was trying to subscribe gemini, now I don’t 😂
Are you using any software to modify your voice? Your voice is very muffled.
I had a mic issue, I’ve fixed it now
Is this Gemini 1.0 or 1.5?
It’s the Gemini ultra 1.0 version that powers advanced. Better than Gemini 1.5
video starts at 3:44
Chat gpt 4 id best 👌
Google most certainly do not value privacy. Their business model is based upon advertisement and that based upon your data.
There is no privacy in the Era of AI!
lol people really can't review things anymore. Says "We are going to compare the individual $20 a month plan vs Gemini" (and not the Teams Plan) but then immediately starts comparing the features found in the teams plan like having the ability to turn off Chat Training and comes to the conclusion that ChatGPT is "winning" privacy smh.
Are you fully awake? You can turn off chat training in the $20 version
I found the video's comparison between GPT-4 and Gemini Ultra a bit unfair since they didn't address the core differences between the models. It feels like they focused too much on specific prompts where one might have an advantage, instead of explaining how each model generally excels in different areas. Also, comparing things like price and storage isn't really relevant to the core AI abilities, making the video feel a bit misleading.
I showed 10 areas that people use most frequently when it comes to LLMs. So it’s a practical comparison for my audience that want to choose between the two.
I feel like not utilizing custom GPT's for all the promts makes this comparison inaccurate, custom GPts make the whole difference, and google lacks this altogether.
Yea I mentioned that several times in the video. I also could have done some things with workspace extension in Google that would be very difficult with ChatGPT though
YOO! WHO ELSE LIKES THIS VIDEO WITH THE HUMAN PILOT IN THE VIDEO!
You made made a 35 min video for simply comparing those two tools and there are no captions to make use of this video...? It could have been done in a less than 3 minutes video damn
I think captions take a little bit of time and this video is very fresh. This is an in-depth review. There are plenty of 3 min I don't show you much - just put the score up and show you 3 examples videos out there if thats your thing...
I’ll add captions. Didn’t realize it didn’t auto generate
Fast forward.
30:34 30:37
you don't know how to give a correct prompt to chatbots. i asked gemini for the same image generaetion and it gave me 6 outputs with only few spelling errors in the first 2 images
you discuss the most advanced AI tools but do not bother to add time steps to your video?
Thumbs up if you prefer Google Gemini!
the secret here is = 2TB
this video could've been 2 minutes long
Oh, Gemini sucks. It's probably the reason why Google is so bad on searching images.
Explain more
What a lot of the AI channels don't understand is that the day that AI actually passes all the tests will be the day the A. I is smaller than humanity. when something is better than 100%, you can't know how much bigger it is
GPT4 appears less racist