Gemini is worse even in premium. Same goes for GPT as i bought premium for it, but chatgpt is just filled with lots of bugs. The chat in between automatically switches to free version even when i paid premium, then i have to start new prompt again, that would mean I have to train it again based on what i was doing.
Really? That's interesting. I've been paying for premium since GPT 3.5 was released, and it always switches to 4o. Every time, I have to switch to 4o mini to use it. Otherwise, the next time I enter, it's on 4o. I haven't experienced any bugs either. It remembers my history quite well, so I never even have to train it. I do have custom GPTs, but I never use them because GPT-4o can figure it out most of the time. I wonder why the quality control hasn't been great. I would definitely report the issue so they can perform more maintenance and updates. I agree with Gemini, but there have been a lot of improvements. It's definitely getting better. I'm waiting for it to at least have more reliable searches.
1:59, Gemini actually got it right. If you read the text it says that it was 13 years ago. 54 - 13 = 41. The statement is still correct, and I think this deserves a point.
Can you explain it in more detail? To me Gemini is definitely wrong. 41 years ago Sally was 13 and her mother was 39. 13x3 equals 39. Making 41 the correct answer.
@@blazer9547 Who cares which is best they all do what 99% of us need. Just use Grok and support Elon and open source. AI should NOT be closed source and manipulated. Claude is closed along with all the other
The upside is that at least Grok gives you the answers you want and doesn't reply with "sorry, that's a sensitive question...blah blah". Sometimes knowledge is what we want to hear, sometimes it isn't. But knowledge is truth. Quit with the censoring.
@@YannR34 i don't see why you would do that!! do for some reason have a problem? i mean its really smart for them to integrate it to X, after all its a good testing bed, and also has millions of users already, which would help even more with training grok. i mean grok 2 is already super good, and whats funny is, it was trained only using oracle data centers...they literally haven't start using the new super computer they built. they move super fast for how young they are compared to other companies, and now they have a super computer online to train their models! so from a model stand point grok 2 seems to be a very good model, and XAI seems to be a great company that moves super fast, like all musk companies...so i don't see why you would say that!
@@BFAO1621😄. I perfectly see why you can't understand it. I have my reasons and I'm not a Musk's fan, actually quite the opposite, and for me it's a reason good enough.
Elon just announced he has invested 100,000 H100 Nvidia GPUs to improve Grok, Grok 3.0 which will be released in December will Be the best model there is. Source: his conversation with Dr Jordan Peterson. for reference, meta used only 15,000 H100 GPUs to train the 405 B model of LLama 3.1. besides Grok has real time access to an overwhelming amount of data through X. Grok is going to be state of the art in just 4 months. don't bet against the man who made rockets a profitable business.
So none are perfect and probably risky to assume any of the answers are correct. Looks like these models still have a ways to go before we can be 100% confident ion using them. Though for some tasks where they just need to use what is on the net and collate it, they could help. I suppose the question hierarchy reaches a point where diminishing returns set in, so if you keep your enquiries below that ceiling then they are solid Assistants. That is mainly how I use them.
I'm an engineering student. And for daily tasks I use Gemini advanced. But for complex mathematical or logical problems, coding etc... Chatgpt 3.5 is decades ahead of Gemini Advanced.
In terms of answering obscure questions and researching things via modern internet 'ability to comb the internet to find specific answers for me' - which is the best?
You are plain wrong. The correct answer is 41 years ago, when the daughter was 13 and the mother was 39 (13 × 3 = 39). If it were 13 years ago as you said, the daughter was 41, while the mother was 67 (41 × 3 = 123).
@@IceSpeed No it’s not. It will have the ability to send prompts to ChatGPT at some point via Siri but the on device features such as summarizing and AI created photo memories that are available now in beta are using Apple’s own LLM, not ChatGPT.
It all come down to integration. It's all well and good to test these things as standalone apps, but until chatgp is integrated into actual useful apps then it will have a very limited userbase. Google are the only ones doing this it seems, but I hate the user experience of it
It's now clear why the Grok version can't be used for free. No one would continue using it after a few seconds of interacting with a product that isn't finished... or rather, hasn't even been properly started.
I wish I could find a way to message you directly. I have been waiting for your video on the Google Pixel 9 series. You're my favorite UA-camr when it comes to anything to do with Google.
Can you please include Microsoft Copilot for sake of completeness? I know it's built on ChatGPT, but it's not entirely the same as OpenAI. Also, Microsoft Copilot is a full-fledged, dedicated mobile app. I would be interested in the results.
Definitely Gemini is much worse, this video is like a Gemini advertisement. For example, while Grok and Chatgpt can create images, Gemini cannot do this, why is this not mentioned in the video?
Is is possible that you have no idea what an AI is good for? What people are using it for 90% is coding, instructions, desition making on complex problems, working with sorting and formating data and of course automatic text generation and correction in any form imaginable. This is what we are all interested in if an AI i capable doing so and not if it understands five different frelling math riddles no sane person cares about or looking up your vacation place.
You try so much to give points to Gemini. The question about the places to visit you only gave the point to Gemini because it show pictures? You should've gave the points for the 3 of them.
u are using fun mode of groq try using regular mode
Gemini is worse even in premium. Same goes for GPT as i bought premium for it, but chatgpt is just filled with lots of bugs. The chat in between automatically switches to free version even when i paid premium, then i have to start new prompt again, that would mean I have to train it again based on what i was doing.
Really? That's interesting. I've been paying for premium since GPT 3.5 was released, and it always switches to 4o. Every time, I have to switch to 4o mini to use it. Otherwise, the next time I enter, it's on 4o. I haven't experienced any bugs either. It remembers my history quite well, so I never even have to train it. I do have custom GPTs, but I never use them because GPT-4o can figure it out most of the time. I wonder why the quality control hasn't been great. I would definitely report the issue so they can perform more maintenance and updates. I agree with Gemini, but there have been a lot of improvements. It's definitely getting better. I'm waiting for it to at least have more reliable searches.
I was surprised to find how poor Gemini Advanced was. It quite often fails at responses and says to just do a Google search.
Chat gpt premium has lost it spark
I have faced the same issue with chat gpt 🔥👍🏽
ChatGPT Premium is amazing what are talking about lol
1:59, Gemini actually got it right. If you read the text it says that it was 13 years ago. 54 - 13 = 41. The statement is still correct, and I think this deserves a point.
It's asking for her age now not 13 years ago. So, if we follow the rules, no point should be given.
Can you explain it in more detail? To me Gemini is definitely wrong. 41 years ago Sally was 13 and her mother was 39. 13x3 equals 39. Making 41 the correct answer.
41 * 3 ≠ 67
Why are you only using fun mode in Grok /😕
Because Elon somehow hurt them and they want to make a review where his products look bad. Rag reporting at its finest.
I really like your comparison. Please, keep making more on the highest available versions.
I use always both ChatGPT and Gemini, and i have to say Gemini is for my use, better and more useful
I want to like Gemini, but Perplexity has been my go-to for a while now.
Basic...
😂 it's a substandard ai, use mistral or Claude
Meaning 😊
@@blazer9547 Who cares which is best they all do what 99% of us need. Just use Grok and support Elon and open source. AI should NOT be closed source and manipulated. Claude is closed along with all the other
Grok 2.0 is out
Chat GPT still better.
@@GamingItbyGameRickster for now
The upside is that at least Grok gives you the answers you want and doesn't reply with "sorry, that's a sensitive question...blah blah". Sometimes knowledge is what we want to hear, sometimes it isn't. But knowledge is truth. Quit with the censoring.
Agreed, I asked GPT and Gemini where I could vote and it said it can't talk about that right now... That was terrifying to me.
now?
Gork can do various tasks depending on the mode. You did everything in fun mode so maybe that is why gork lacked
WoW, it's cool to see our country get mentioned in your video. Love your content! 🔥
Greetings from Baku 😎👋
I love my Pixel Fold with Gemini works so nice
need claude AI comprasion too! It's getting ahead of many competitors!
True, 3.5 sonnet is awesome although I haven't payed for premium. If you have premium how has it been for you?
@@micahjohnson9184 I didn't pay either.
Closed source ai is a threat to humanity. Use Grok
I didn't even know the existence of Grok 😅
Yep and I think that I'm going to keep it that way.
@@YannR34 i don't see why you would do that!!
do for some reason have a problem?
i mean its really smart for them to integrate it to X, after all its a good testing bed, and also has millions of users already, which would help even more with training grok.
i mean grok 2 is already super good, and whats funny is, it was trained only using oracle data centers...they literally haven't start using the new super computer they built.
they move super fast for how young they are compared to other companies, and now they have a super computer online to train their models!
so from a model stand point grok 2 seems to be a very good model, and XAI seems to be a great company that moves super fast, like all musk companies...so i don't see why you would say that!
@@BFAO1621😄. I perfectly see why you can't understand it. I have my reasons and I'm not a Musk's fan, actually quite the opposite, and for me it's a reason good enough.
Unlike ChatGPT
Now do Grok2
Does grook2 have images and documents upload to prompt?
Thank you for including Grok
I tried the book summary prompt as it is exactly in the free version of Gemini, unironically was able to provide it lmao
Elon just announced he has invested 100,000 H100 Nvidia GPUs to improve Grok, Grok 3.0 which will be released in December will Be the best model there is. Source: his conversation with Dr Jordan Peterson.
for reference, meta used only 15,000 H100 GPUs to train the 405 B model of LLama 3.1.
besides Grok has real time access to an overwhelming amount of data through X.
Grok is going to be state of the art in just 4 months.
don't bet against the man who made rockets a profitable business.
and its uncensored.
😂
Grok 2.0 released soon after this video too, the update is great + the image generation. Poor timing to post the video tbh.
Hello from Georgia👋It was surprising to hear you mention it.
You have good food.
@@bolle666 Thanks 🙏 It’s true, we have variety of good food, especially Khinkali and Khachapuri
Gemini is my favorite! ♥
Google shill.
I love both gemi or chpt
@@RealtyWebDesignersyep gemini just pushes propaganda
I eagerly await the day when an AI will actually be helpful, consistent, and reliable.
Grok is gaining more computing power than ChatGPT and Gemini combined so pretty soon not only will it catch up to them but surpass them as well.
Iooking forward to this. Can't wait
Gemini actually answered the sally's mother question correctly which is 13 years ago. i.e. when she was 41 years.
it's wrong, the correct answer is 41 years ago, when she was 13
So none are perfect and probably risky to assume any of the answers are correct. Looks like these models still have a ways to go before we can be 100% confident ion using them. Though for some tasks where they just need to use what is on the net and collate it, they could help.
I suppose the question hierarchy reaches a point where diminishing returns set in, so if you keep your enquiries below that ceiling then they are solid Assistants. That is mainly how I use them.
Such a great comparison. This helps me make my decision to go with ChatGPT
I'm an engineering student. And for daily tasks I use Gemini advanced. But for complex mathematical or logical problems, coding etc... Chatgpt 3.5 is decades ahead of Gemini Advanced.
Thank you and please do more of these videos.
Points should be removed if they hallucinating na? 😅
In terms of answering obscure questions and researching things via modern internet 'ability to comb the internet to find specific answers for me' - which is the best?
Gemini got the Sally question right. 13 years ago was the answer. 54-13=41.
You are plain wrong.
The correct answer is 41 years ago, when the daughter was 13 and the mother was 39 (13 × 3 = 39).
If it were 13 years ago as you said, the daughter was 41, while the mother was 67 (41 × 3 = 123).
@@waldamyInteresting
Azerbaijan. I didn't expect it 🇦🇿❤
You are using "Fun mode" in Grok.. donno if that affects the answers?
Do a Gemini vs Apple intelligence video
Apple Intelligence is powered by ChatGPT
@@IceSpeed No it’s not. It will have the ability to send prompts to ChatGPT at some point via Siri but the on device features such as summarizing and AI created photo memories that are available now in beta are using Apple’s own LLM, not ChatGPT.
He will once it gets fully released
Grok is a terrible name.
Look at Elon's kids name. He isn't known to name things well
@@NeptuneSega agreed. X is also an atrocity
Sounds like a frog
Worse than llama, Chat GPT etc? To Grok means to think, come to a realisation.
@sntki8561 Yes, but if you do not know the meaning, it sounds like a frog screaming
Love from Georgia 🇬🇪
It all come down to integration. It's all well and good to test these things as standalone apps, but until chatgp is integrated into actual useful apps then it will have a very limited userbase. Google are the only ones doing this it seems, but I hate the user experience of it
❤ best 🎉
Gemini !!!! 🗣
It's now clear why the Grok version can't be used for free. No one would continue using it after a few seconds of interacting with a product that isn't finished... or rather, hasn't even been properly started.
What about copilot
I wish I could find a way to message you directly. I have been waiting for your video on the Google Pixel 9 series. You're my favorite UA-camr when it comes to anything to do with Google.
Can you please include Microsoft Copilot for sake of completeness? I know it's built on ChatGPT, but it's not entirely the same as OpenAI. Also, Microsoft Copilot is a full-fledged, dedicated mobile app. I would be interested in the results.
I'm expecting a video from you related to video boost in Pixel 9 Pro with 8K upscaling
Do a ChatGPT o1 Gemini advanced and Grok 2
claude 3.5 sonnet is BEST I Try with your all Q\A = result 100%
So the Grok logo is the same as the Deutsche Bank logo?
Chatgpt is best one 💡
Where is pixel 9 series review?
You're using it as a 'search engine'; what matters is new content creation.
Grok is worse because xAI (the company being grok) was founded most recently, and training these LLMs takes time.
Definitely Gemini is much worse, this video is like a Gemini advertisement. For example, while Grok and Chatgpt can create images, Gemini cannot do this, why is this not mentioned in the video?
Because of grok
grok is not so old like others
Try the new gemini live
Chat gpt is biased intelligence . Im moderate but don’t like how chat gtp is super woke based
You’re fired I’m afraid!
Is is possible that you have no idea what an AI is good for? What people are using it for 90% is coding, instructions, desition making on complex problems, working with sorting and formating data and of course automatic text generation and correction in any form imaginable. This is what we are all interested in if an AI i capable doing so and not if it understands five different frelling math riddles no sane person cares about or looking up your vacation place.
Leaving meta ai out of this is crazy
Gemini is an embarrassment for everybody who's ever worked with coding
You forgot llama
You try so much to give points to Gemini. The question about the places to visit you only gave the point to Gemini because it show pictures? You should've gave the points for the 3 of them.
I use gemini
Already lost.
I'm pretty sure nobody uses Gemini because of how bad it is.
like if you agree
I use Gemini all the time so it's almost nobody at the very least plus the video doesn't support your argument
I use it and it’s good.
gemini was bad. now it's actually really solid. better than chatgpt in many cases.
I found ChatGPT to be the most mature among all AI assistants.
@@danielayanfe3592 u need to try gpt 4o
Must name it Groot😂