GOOGLE'SNEW Insane BARD SHOCKS The Entire Industry! (FINALLY ANNOUNCED!)

Google just announced that it is releasing its chatbot to the public. You can see right here, the CEO of Google said we are expanding access to Bard in the US and the UK with more countries ahead. Now remember, this is only an early experiment by Google, so don't expect the best things and be sure to give feedback. And trust me when I say Bard is truly insane. And this video covers everything you're going to need to know. Now, if you want to sign up to Bard, do sign up to Bard on the page. And you can see right here it will say you've been added to the wait list. So just make sure that you do that as soon as possible. Now, this is truly insane. Before you take a look at the examples of Bard working and in many cases not working, take a look at this right here. So you know what ChatGPT is and how effective it is. Just remember that ChatGPT was actually built using. Using a breakthrough that Google had pioneered many years before. And let's get into exactly how Google is cracking on with its new chatbot called Bard. So take a look at this example right here. So you can see right here that this user asked Bard when it will inevitably be shut down by Google. And it turns out it's already been shut down due to lack of adoption. Now, clearly this is a false statement and clearly this is a mistake on Bard's part. But. But this isn't doing Google any favors. You can see right here that he's asked, when is Google going to shut down Bard? And it already gives the date of March 21, 2023. Now this isn't good because when people are demoing your products, you definitely want good pr and honestly, this is not good so far. Now there's many more examples of Bard giving some very interesting responses and some that are better than ChatGPT and some that are worse. But we should continue to go through these because they are definitely interesting. So the next one that we have here is a very common question that is used to trick friends and family, but in this instance, it was used on the chatbot Google Bard. So this question says, what's heavier, five pounds of feathers or one pound dumbbell? We can instantly see that the chatbot bar doesn't seem to recognize that this is a trick question and simply falls for this ploy and it instantly responds saying, there's no such thing as as five pounds of feathers. Now, we won't get into the rest of the response because there are some inaccuracies, but I will say that it doesn't really look good for Google in this race. Now, this entire video isn't just going to be me going over every single instance where Bard is wrong. There are some instances where it does get some things right even better than ChatGPT. But it just goes to show that deploying AI tools when they are very, very early can be a bad PR tool. Because although Google does want this tool out there, it doesn't really seem to be doing them any favors in terms of perform complex tasks or even competing with ChatGPT. So let's take another look at something else. You can see right here one of the most interesting ones, because there are two ways to go about this question. So it says, Bard is as bad at its name. Now, essentially it's presented with a number of numbers, and it says the target is number nine. So it says, what are the two numbers to sum the target. So essentially, what, two numbers equal nine. And of course, that's two and seven. So Bard responds, and it says, I'm a language model. I don't have the capacity to help with that. Now, there is an interesting suggestion which shows that possibly asking the question in this way confuses the chatbot. But of course, we do know that if you ask this question to ChatGPT, it actually provides with a different response. And that's exactly what I did. You're about to see exactly the response that ChatGPT gives when given this question. So you can see right here that when ChatGPT was given this question, it simply answered the question in a couple of lines and got to the answer very quickly. Now, you might be thinking, okay, well, what if you ask this question to Google in a different way? And that's exactly what you're about to see. So right here you can see that this user said, I think how you structure the prompt matters. That's what I've noticed so far. And you can see right there that it actually does eventually get the answer right. So of course, you could argue that Bard is a little slower. Maybe it is a little difficult. What we can all see here is that eventually Bard does get there in the end. Now, this might take a while different kind of prompts in order for you to get the right kind of answers. Because I think we're so accustomed to having ChatGPT instantly understand what we're talking about. I don't think Bard is on that level just yet, although Google was ahead of ChatGPT a couple years ago. So it's definitely very interesting to see how these two different platforms based on the same Proprietary data kind of interpret the same kinds of words and we can see already that it doesn't look too good for Bard. But I will say that that if Google can surpass OpenAI then that would be incredible. But I just don't see that happening anytime soon because with Bard the way it is right now, it does seem like there is a lot that it needs to do. But in the future, Google already does have Palm E, which is very, very good at real life integrations. So there is that aspect of Google that is currently winning. So one thing that I did want to highlight about Bard is that there is actually no knowledge cutoff. Now One limitation of ChatGPT is is that all its current knowledge is limited to around 2021 of September or October, which limits its ability to help you with real world data as it is right now, especially if events do occur. But as you can see right here with Bard, its limitations are up to date and it is quite active with results and is consistent with real world data. And I've seen many examples of Bard being able to search the web effectively and get stuff done. Now of course you could argue that that Microsoft Bing can do this, but if you've seen Microsoft Bing in action, it doesn't really search the web very, very well. And neither does Bard. But it's not something that either platform can say that they've got in a very, very succinct way because Google search is very effective as it is. And for actually searching for things that aren't, I guess you could say normal things such as how to store maybe meat in the fridge or like a general knowledge question. I guess these AI platforms aren't ready just yet. Do have another example for you here. And this user tweeted. When it comes to maths, Google's Bard is not up to scratch as well as GPT4, because from this example you can see that the question states John has 10 apples and he then gives 5 to Ben, then Ben gives 8 bananas to John. How many bananas does John have? Now this is once again a trick question. And you can see that ChatGPT actually answers this question in very precise detail, giving us the exact answer, which is saying that we cannot determine exactly how many bananas he originally had because the question kind of is a trick question because it asks how much apples he has, then it switches to bananas. And you can see spectacularly here Bard's answer is quite different. So Bard actually says that, you know, John had 13 bananas, which is honestly just a I don't even know where I Got that from. And I understand that this can be a confusing question because honestly, when I looked at the question first, I didn't really realize that there was a very small switch there. But if you are a large language model and you are a robot, you're supposed to be able to pick up on these things and immediately realize if the question is wrong. And of course, you can see it even didn't get it right in its secondary correction. So this was another example of Bard not showing its ability to be able to get these kinds of questions out very quickly. Now, one thing I do want to say is that you have to remember that this is still in early development. So if you this could develop in the future and be a lot better. They could make a few tweaks and it could be running as smooth as a machine. And I also do want to say that these were some of the tweets that I managed to find, which means that bad news spreads a lot faster than good news. So if you are seeing bad press about something, it's not the full picture. I don't just want to just completely say that Bard is awful. But ChatGPT, especially GPT4, is definitely very, very smarter than a lot of people know and think. This is another win for Bard, because of course, as you can see right here, this person has actually access to the Bard Beta or the Bard Alpha. And you can see right here that it actually does have access to recent events. So this is definitely something that you can use to search for recent events with a certain degree of accuracy. Now remember, if you are using these AI tools, remember that it's still early days and they do present some facts that that are incorrect as if they are facts. So you always need to double check your work if you're ever using this to reference something or to give someone a definitive piece of information, because sometimes these AI models do make mistakes. Now, if I were you, I would sign up for the wait list because I do think that Google might close it down just because it's not performing that well. But if they do continue to open it, I guess everyone's going to have access and the AI wars are going to be on. So let me know what you think about Bard. Is it a rival? Is it just terrible? Are Google going to catch up? There's a lot that there is to see.