China is building a 1 km wide space station with solar panels on top of it. Electricity will be generated and sent to the earth. This bike you are seeing is real. It’s not an AR video. A company has invented a bike which can run at a speed of 200 kmph. If we talk about China, they are working on a robot test control. In South Korea, a robot has been made which is in liquid form and can go through metals. Because of Neuralink, a person who is in a paralytic state can speak in his own voice. AI is also coming in Netflix. We can watch movies by adding prompts like chat and GPT. Hi guys, my name is Sahil Khanna and let’s start today’s technique. The first news is about Netflix. AI is also coming in Netflix. In the starting, it will come in iPhones. After that, it will be seen in other devices as well. Here, you can write a prompt in the search tool as to what type of content you want to watch. You will be recommended that type of content. A layer of chat GPT is being used. This is a go-to process where you can take your current feature app to the next level. The chat GPT wants FDA approval for its CDR GPT which will speed up the process of drug evaluation and research. It will take years for AI to be seen in real-time. AI will be able to review documents, recognize patterns, and analyze data in real-time. AI is helping in making medicines and healthcare. Sam Altman has said that the US is leading in AI race, whether it is OpenAI or Google. But there is no guarantee that this lead will continue. The US government is imposing sanctions on China. This might create a necessity. Inventions are made out of necessity. It might not be that you put so much pressure on China that it will get used to making its own products. The simple thing is that if I keep giving you everything, you won’t feel like working hard. You will be sustained at one level. When there is pressure in life, only those people come. Like they say, coal is converted into diamond under a lot of pressure. It is the same with humans. Sam Altman talked about Donald Trump’s stance. In the beginning, he was very supportive of AI. But now, he has become very cautious. AI progress can stop because of this. A research has been conducted which has shown that if you accept smaller answers than AI, you will see more hallucinations. For those who don’t know, there is a big problem with AI. If we ask it, it starts making mental stories. These mental stories are called hallucinations. If you ask for smaller answers, you will see more hallucinations. I will show you a chart. If we talk about natural instructions and short answers, there is a big difference in hallucinations. If we talk about small models like Gemini 2, Flash, Quint, or Gamma, there is no drastic difference. But there is a big difference with big models. If we look at the sonnet of the cloud, it is giving the best response in hallucinations if we look at it regularly. There is a big concern for developers. If you are training your own AI model, we are also training our own model, the bigger the answers, the bigger the cost. We have used a layer at the back. We are not paying for our own layer. We are paying for the back layer. Ideally, we will try to get at least one answer in each word. Because we have to pay for each word. So, we put a system prompt. We have to give a prompt for how to behave. So, we have to put a system prompt. Please do concise answer. So, if you are using a concise word, there is a hallucination. This means that developers have to increase their expenses to reduce the hallucination. The new model of Mistral has outperformed Lama 4. Cloud’s 3.7 solid model is around 90% better than that. And the cost is very less. Context window is 8,28,000. And it can run on a 4GB server. It can be run in an office in less power. Which is a good thing. If you see the chart, you will understand where is Mistral’s medium 3. Lama 4’s maverick is in almost all. In multimodal, I will say that in some tests, Lama is winning. In knowledge, it is winning. But it is not winning anywhere. This model is stronger than 4.0. If we talk about coding, you can see how it is competing with other models. It has lost to Sonnet. It has lost to DeepSeek. It is competing with 4.0 in 50-50. It has won against Command A. And it has defeated Lama. I would like to tell you one thing which many people don’t understand. A company produces many models. If we talk about 4.0, you have the option to change the model. You can use 4.0. You have different models like O3, O3 mini, O3 high. There are many options. Why are they giving so many options? Even in Gemini, you have the option to use all the models. Why are they giving the option of multiple models? When a company produces a model, it is trained for a specific task. It is possible that Gemini’s 2.5.0 version which is the latest, is good in coding. But it has lost to the old version in some task. Let me tell you which models are good for which task. This is the website of OpenAI. It has given a clear cut. If we talk about everyday tasks, 4.0 is very good for brainstorming, summarizing, etc. You should spend your maximum time on 4.0. It will be 4.1 soon. It will perform all the tasks. If we talk about 4.5, it is good for creative tasks. It has better emotional intelligence. If you change the dropdown, you can see that this model is good for a specific thing. 4.5 will stop it. But 4.5 is good for writing. If we talk about 4.0 mini, it is good for visual reasoning, programming, or quick queries. It is good for summarizing. If you use a chat GPT model, it is good for normal commercial apps. You don’t have to go to 4.0. But if you want a basic understanding, like we use our MIDI model for our large language model. The back layer is of MINI. If you want to know something, we give it to them in a better way so that they can understand. If they structure it according to us, they can give a better answer. This is how we are using it. If you are asking why I am not using the latest model, it is because of the cost. We have to explain what we are doing. The calculation is done in our server. We don’t have to buy a dangerous model and spend a lot of money. If we talk about O4 mini high, it is good for complex tasks like carrying SQL, data extraction, solving math problems. O3 model is one of the most powerful models in advanced reasoning. You can ask cross-questions here. Normally, it is recommended in 4.0. But if you are talking about a complex task and you want to discuss it with a document, it is good for that. If we talk about strategic planning or detailed analysis or coding, it is good for all of them. O1 pro model is accurate in doing complex tasks. It is very good. I would recommend if you are using chatGPT in a regular phone, use O3 model or use 4.0 when it becomes 4.1. 4.0 and 4.1 are very light. They give a quick reply. It takes some time. Most of the work is done in O3 models. Before moving ahead, I would like to remind you to hit the like button, subscribe to the channel and share as much as you can. If we talk about mystery people, their revenue has increased 3 times in the last 100 days. In February, they launched their enterprise version of Leech AdBot which is their open source. It has integration with Microsoft SharePoint, Google Drive, Content Management Systems. Recently, they have raised 6 billion dollars on valuation. Business is also growing. Can it become the third AI company? Which are the first two? I would say fourth. Let’s talk about valuation which is the most important. I think X their model is growing better because people are using it. They are getting live data on Twitter. Entropic has proved itself on many levels. In coding and modding. Google is also there. It can be better than Lama We can say that it can be better in open source race Because of Neuralink chip, a person who is in paralysis His whole body is not working He can speak in his own voice This is a little complex, let me explain how it is working The jeep is working According to the movement of the jeep The mouse is working and it is able to work with the mouse He has to press his jaws, which means he is clicking And the jeep is moving, which means it is tracking It means his eyeball is working and the jeep is working The brain is checking the signal, how the jeep is moving Based on that, the mouse is being used and it is working This is the magic of AI To understand this, watch this video You will understand how it is working It gives word suggestions for the situation And it pushes apps out of the way I do a lot of different things with keyboard shortcuts So I created a keypad using the Mac accessibility keyboard To make some keystrokes easy to use Things like select all, copy, paste, undo And ways to navigate the page are all really useful And I can use the toolbar and some shortcuts Scientists have discovered a process Which can repair or protect the microchondria Because the microchondria gets damaged, there are a lot of problems You must have heard of Alzheimer’s Which is very common It can be cured Or precautions can be taken So that the microchondria doesn’t get damaged Apart from Alzheimer’s, talk about Parkinson’s Talk about ALS, type 2 diabetes Or accelerated aging It can be stopped It is not that you will become immortal And you will not become old If microchondria is controlled But it will be beneficial for those who have microchondria Microchondria is the powerhouse of the cell It gives energy to the cell So whatever treatment we are doing We will feel more energetic There is a new AI It is called Face Age You will have to take a selfie You will know the chances of cancer survival Basically, it is not for everyone Doctors will take it out Patients will be tested A model train has been done People’s photos will be taken People who are actually suffering from cancer They will be taken photos It will be seen with the face What is the difference in their age How fast they are getting impacted Based on that prediction AI will be able to solve the problem I say Cures should come soon It will be fun Cures should be applied It is not like We have done a research At scale The disease should be over Baidu has filed a patent Animal voices can be translated It is too early We are getting a lot of criticism Do something I am wearing earphones It has AI My dog is barking Give me food I am telling him He is able to understand It will happen We are filing a patent It is too early We will see it happening soon It will be amazing People were keeping pets So that they can become partners They are able to communicate They are not able to communicate But after this device They will become partners Scientists have developed a liquid robot Which can change its shape Watch the video What is going on It came And left Watch it again Basically Its movement is controlled by ultrasound waves And it can move It is a good thing It can be used for drug delivery If you want to send something It will go with the medicine Then it will be pulled back Or a small hole Can be made Or disaster zones It can provide supplies Technology is doing a good job Chinese farmers are using a robot For pest control The robot has sensors Cameras It can detect pests It can give treatment accordingly Pesticides will be less Environment will be better If we talk about Apple They are developing a new chip For their smart glasses It can take 2 years Apple glasses are not going to come in 2 years I think the world will go far I think The world of glasses Is going to be achieved very soon Meta has launched ribbon glasses Which is very successful Google has showcased They have showcased their operating system Which will work in AR We live in android world We have to use phone Meta said They will make a device Which will replace the phone It has not happened yet But the plan is It will happen very soon Because when AR glasses were there Technology was not at that level Today we have AI And AI Is at a very low level Which can be run locally Today I can tell Meta Please do this He is listening to the task Executing it He has a camera He can listen and discuss Earlier it was not possible Today internet is very fast Chips are very fast Very low power Battery is also changing So somewhere Glasses are going in a very successful direction Now Who will win this race This is a big question Meta has an advantage They have their own large language model Which can be run in the phone They will run all the apps All tasks will be done here But you will say Google has its own operating system Google has its own app network So all the apps will be with Google They will port it in a minute But it is not like that If I talk about glasses I will not like to watch movies What will be the requirement in glasses Chat, message, call Navigation These things will happen in day to day tasks Please call this number He will show you Are you talking about this These things will happen Don’t expect We will watch it on Netflix VR devices are better for that In VR game Meta faced a problem Because in VR It was difficult to make an ecosystem I don’t think Meta will face the same problem in AR Or any company Because there are many open source large language models Which can be integrated with glass Glass will be used as a hardware If we talk about Apple They are going to do something About multiple cameras I heard They will put some sensors in the airpods Maybe we will get to see cameras in the airpods I don’t want cameras in the airpods It should be glass only Use glass only If multiple cameras are installed You will be able to take a lot of dimension data Let’s talk about new glasses There is going to be a super sensing feature Where it will remember the previous And it will remind you Where is my key It is here It will do this You can say It is a privacy concern It is recording everything It is able to discuss Battery consumption is also increasing In the trial Battery is running for 30 minutes It is a live AI feature In the next level of devices They plan to take it for 1 hour Which device will it be? The question is There are a lot of devices in the Metaglass They are saying they will launch 2 devices next year I was hearing There is going to be another device this year What will happen? Time will tell I am waiting for new devices If it is not coming Let me know I will buy the old one I can wait I want to buy a lot I have used some I don’t like the fit I won’t be able to use it It will be the same Like I bought my Meta VR devices They are lying there They are not comfortable I have heard They are coming in different fit I will have to buy it physically It is not available in India They have revealed They have revealed a bike Which looks like Star Wars You can go at a speed of 200 kmph It looks fun What are the other details? How long will it run? Is it practically viable? There are a lot of questions Microsoft has banned its employees From using DeepSeek Because of privacy You can use DeepSeek in Microsoft’s Azure You can use DeepSeek in Microsoft’s Azure Some will say it is hypocrisy But the story is If you use it in Microsoft’s server Because DeepSeek is an open source So you are running it in your server, data is not going to China, this is what Microsoft has to say. Huawei can launch their laptop on May 19, this is a big change with HarmonyOS. They had a contract with Windows that was about to end, now they have to do this forcefully because there is a conflict between the US and China. HarmonyOS will have to come, but there is an advantage for them, it looks like an ecosystem is being made. HarmonyOS is in the laptop, in the phone, they have all the devices. If you see, they have a watch, they have a phone, they have a TV, they have tablets, basically all the devices, so everything will be connected to each other, it will be an ecosystem game. And the Chinese people use Huawei, it is a good big audience segment, so if an ecosystem is made there, then somewhere other companies will also suffer a loss. I was talking about Necessity in the beginning, I can see it there, it has been invented from Necessity. Huawei, which seems to be growing, has grown out of helplessness, today it is able to make its own AI processors, today it is able to bring a lot of innovations in its phone. The Chinese are making a solar space station, which is going to be 1 km wide, it will be above 36,000 km earth, there will be a solar panel, it will capture energy, it will be converted into a microwave, there will be a receiver inside the earth, it will be sent, and electricity will be generated from there. If you are having a deja vu, then we had covered this news recently, but at that time we were talking about lighting a bulb of less power. Here we are talking about deploying a 1 MW station by 2030, 10 MW by 2035 and 2 GW commercial plant by 2050. It will be a clean energy, there is no tension of electricity, there is no pollution, all electricity is coming from space. My question is, if someone comes out from the bottom of the microwave beam, it will be finished or not, it will be made into a no-fly zone, and it can also be like a bomb. If you are shooting a beam of 1 MW electricity, it may come slow, but it is going to be very hot, if you want to finish something, then point the satellite there and kill it, it will become a weapon. How many people agree with me, tell me in the comment section, is this weaponization or not, it is being said that there is a solar panel, but is there a way to make it a weapon? China is making the world’s largest man-made forest, it will be completed by 2050, it was started in 1978, it is 4828 km long and 1448 km wide. Basically, it is being made so that one section is stopped, the storm will be saved from there, they are making their logistics, along with that the pollution is also being improved, clean air will be available. The US has launched a drone that is using AI and can pick up its targets, its speed is going to be 160 km per hour, its name is Dread Dragon, it can travel up to 400 km, it has a weight of 20 kg, and it will be launched in 10 minutes, it can also take explosives. Rest, the environment in the world is made of war, so it can be used somewhere. Rest, AI is also coming in Google Chrome, where you will be saved from scams and shady sites, it will run locally, I was telling you earlier, it is not happening through the cloud, here they will use Gemini Nano model, which is going to be AI-driven, it will run locally and will be able to save a lot of things, it will also come in Chrome browser phone, it will also come in the laptop, it is a great thing, now you have the option, if you want to disable it, you can disable it, but it will help you. They say that the scams where there is impulsionation, there can be 80% drop, the fake airline customer frauds, there will also be a lot of control, this scam runs where there is a fake number of support, there are a lot of scammers, what they do is, let’s say they want to hack a US person on Google, so there is a crypto app there, they rank their website on its support number, so someone searches, if there is a problem in someone’s website, they search for its support, the support is theirs, of the fraud. The fraudsters get a call, the fraudsters talk, and they say that you show your details, we will take their access and remove all the stuff, so such frauds run, all these frauds will be seen to be stopped because of AI, it is also helping people on the right path, you can also say that because of AI, wrong work is also seen to be increasing. Rest, Google Maps can now scan with the screenshot of the iPhone and save the location, this had to happen today or it had to happen yesterday, now Google has so much knowledge that where is this photo from, based on that it can do everything, it is a good thing, if I have seen something on Instagram, go to this place, click that photo and give it to me, and it has saved that thing with me for the long term, it is good. Rest, Aridas has launched 3D printed sneakers, and it is fully printed, earlier they had launched one, it was just one part of the 3D printers, in this, it is completely 3D printed, it is a good thing, shoes are being launched, send the 3D printed model at home, we will launch it at home, we just need the material, we will take a printer worth 15,000 and we will print shoes in it. Rest, there is going to be a lattice design, where you will have the floor of air, water, debris, meaning there will be no dust inside, and air will keep coming in, it is priced at $140, it is available in selected places, if someone has bought it, tell me how it is, rest, with this, today’s tech news ends, I hope you enjoyed it, like it, share it, spread the videos, they are doing a great job, I know, rest, you also know, so only if you share, the world will know, only then our channel will see growth, we have been seeing the views stopping for some time, maybe because of the situation that has been created, but if you don’t spread it, how will it work, if the views stop, my motivation ends, to be honest, now I have motivation, I am trying to maintain it, increase the views, share it, share it on twitter, this is called AI tech news, this is called tech news, spread it, create air, make it real, spread it, spread it, only then the views will increase, I am doing it honestly, our data points are good, retention is good, people are watching till the end, supporting, but not sharing, not subscribing, so bye, good morning, good afternoon, whenever you are watching the video, Jai Hind, Vande Mataram, remember me in your prayers.