00:00So, Microsoft Bing just rolled out a new feature called Deep Search, which is an AI-powered upgrade that's changing how we search online, and it's honestly amazing.
00:09Deep Search is a new thing from Microsoft, launched on December 5th, 2023.
00:14It's still being tested by a few users, and what's cool about it is that it uses GPT-4 to give more detailed answers to your searches.
00:21It's great at understanding complicated or unclear questions that normal search engines can't handle, and it gives you way more specific and relevant answers.
00:28For instance, like if you were searching for loyalty card programs in Japan, normally you get basic stuff like Wikipedia links or blog posts.
00:36But with Deep Search and Bing, you get a whole new level of detail.
00:40You'll see a list of popular loyalty cards in Japan with their logos, descriptions, benefits, and website links.
00:46There's even a comparison table, a map showing where the stores are, and a graph of their market shares.
00:51All this info comes from Deep Search using GPT-4, based on your search and web data.
00:56Now, how does Deep Search work with Bing?
00:58It's not replacing Bing's usual search, but it's an added feature that makes web searches richer.
01:04It combines GPT-4's AI capabilities with Bing's search tech to filter and sort results by how relevant, detailed, credible, and recent they are.
01:13Deep Search is a mix of Bing Search and GPT-4's AI, creating a unique and powerful search experience.
01:20What's really interesting is how you can search with Deep Search.
01:23It's not just typing keywords.
01:25You can use natural language, pictures, or even voice.
01:29You can ask questions, upload images to get info about them, or use voice commands.
01:34It can even handle multiple types of input at once, like text and image.
01:38For example, upload a dog's photo and ask what breed it is, and Deep Search will tell you.
01:43It's a game-changer in searching the web, making it more user-friendly and interactive.
01:48Microsoft actually showed a GIF of Deep Search on Bing, and it looks awesome.
01:53There's a button to switch between regular and Deep Search.
01:55The search box has icons for text, image, and voice inputs.
01:59There's a progress bar showing how long the results will take, which depends on your query's complexity.
02:04The results come in a cool, interactive format, with tabs, tables, maps, graphs, and more.
02:10You can also filter and refine your results.
02:12Now, when considering how accurate and good the results are,
02:16Deep Search sorts them by looking at how closely they match what you're looking for,
02:19how much detail they have, how trustworthy they are, and how recent they are.
02:23It shows quality indicators like ratings or verification badges, and uses human feedback to keep improving.
02:30However, there are a couple of problems with it.
02:32Firstly, it's slower than usual search engines because it uses GPT-4 to get answers, which can take around 30 seconds.
02:39Also, not everyone can use it yet.
02:41It's only available to certain groups like people at Microsoft, researchers, and some partners,
02:47and there's no information about when it will be open to everyone or if it will be free.
02:51Now, how does Deep Search change things for users and SEO?
02:55That's search engine optimization.
02:57For regular folks searching online, Deep Search could be really helpful.
03:01It can find answers to tough questions that normal searches might miss, and it's kind of fun to use in different ways.
03:07But for people who run websites, it's a bit tricky.
03:10Their web pages might not show up as much in search results because Deep Search gives straight answers.
03:15It could also mean they have to change how they make their websites more visible in Bing searches.
03:20However, it's not all bad.
03:22There might be new ways to improve their sites for these searches.
03:25Microsoft describes Deep Search as a fresh way to use AI to explore the web.
03:30It's not going to replace Bing, but it's an extra feature.
03:33They're still improving it, listening to what users say.
03:36This is part of Microsoft's big push to make AI useful for everyone.
03:39So, what do you think about Deep Search?
03:41It seems like a big thing for search engines and AI, doesn't it?
03:45Share your thoughts in the comments.
03:46Now, let's shift our focus to something really intriguing that's been happening behind all the news about new AI tools like Google's Gemini and others.
03:55An AI alliance is forming.
03:57Yes, some of the biggest names in the AI industry are coming together for this initiative to drive AI innovation in a more open and responsible manner.
04:05This is a significant move and could greatly influence the future of AI and its impact on our society.
04:11All right, so, the AI alliance, announced on December 6, 2023 by IBM and Meta, is a group of more than 50 different companies, universities, charities, and other groups.
04:22They're all coming together to improve how AI is made and used, making sure it's safe and good for everyone.
04:28They also want to keep an eye out for any risks that might come up.
04:31IBM and Meta shared that the alliance has four main goals.
04:35Ensuring AI's safety.
04:37Encouraging groups to work together and share ideas.
04:39Bringing in a variety of viewpoints into AI development and making sure AI benefits everyone in society and the economy.
04:47They're also planning to work with governments and other important groups to make sure AI is used in a way that's ethical and meets set standards.
04:54Now, the AI alliance is a big group with many important companies from different fields like AMD, Dell Technologies, Red Hat, Sony Group, Hugging Face, Stability AI, Oracle, and the Linux Foundation.
05:07Each company brings something special to the group.
05:09For example, AMD is great at high-performance computing, Dell Technologies has lots of IT solutions, Red Hat is known for open-source software, Sony Group is a big name in entertainment, Hugging Face specializes in natural language processing, Stability AI focuses on making AI fairer, Oracle is a top player in cloud computing and databases, and the Linux Foundation supports projects that are open-source.
05:33Together, these members spend over $80 billion every year on research and development.
05:38That's a huge amount of money.
05:40With this kind of budget and power, they can create amazing AI systems that could help solve big problems in areas like healthcare, education, and more.
05:49So, how's this AI alliance going to work?
05:52They've set up a governing board and a technical oversight committee.
05:55The governing board will have people from each member company, and they'll decide the main goals and plans for the alliance.
06:01The technical oversight committee is made up of experts from different areas, and they'll handle the technical side of things, like setting standards and best practices.
06:09The alliance is also planning to work with other AI groups and initiatives.
06:13One big part of the AI alliance is working with academic institutions.
06:17Places like CERN, NASA, Cleveland Clinic, and various universities will be involved.
06:22They'll offer scientific advice and resources, like data and tools, and help with projects that make AI more accessible and collaborative.
06:30Now let's talk about Meta, one of the main leaders of this alliance.
06:34Meta is known for supporting open-source AI models and responsible development.
06:39They've released big AI models for public use and developed platforms and tools to help people work with these models.
06:46They've also been working on making AI safe.
06:48But Meta has made some controversial decisions, too, like dissolving their responsible AI team.
06:54This team was focused on making sure AI models are built ethically.
06:58Meta says this decision was part of reorganizing things and that the team members have moved to other AI-related teams.
07:05They still say they're committed to responsible AI development, and being part of the AI alliance is a step in that direction.
07:12However, some people are skeptical about Meta's moves.
07:15They point to past issues with Meta's AI systems, like privacy concerns, misinformation, and bias.
07:20They're calling for more careful monitoring of Meta's AI activities.
07:24Lastly, it's worth noting that some of the biggest AI developers aren't in the AI alliance.
07:29Companies like Microsoft, Google, OpenAI, and Anthropic have their own projects and have formed a different group, the Frontier Model Forum.
07:38This group is focused on responsibly developing large-scale generative models, which are really advanced AI that can create realistic content like text and images.
07:47So, what do you think about the AI alliance?
07:49Will it be helpful?
07:51Do you trust Meta and IBM to lead it?
07:53Let me know in the comments, and don't forget to like this video and subscribe for more AI and tech updates.
07:58Thanks for watching, and I'll see you in the next one.
Comments