音声録音の一部の文字起こしを生成します。
|
モデルにクエリを直接実行して、返された結果をテストできます。その場合、Cloud コンソールでさまざまなパラメータ値を使用することも、Vertex AI API を直接呼び出すこともできます。
|
|||||||||||
自由形式のプロンプト
音声文字変換を行います。音声のみを抽出し、背景の音は無視します。
レスポンス
Every moment around the world, people and organizations are generating data that can be extraordinarily useful, and I think we have to find a way to harness that to solve problems. The challenge is been a lot of these data sets are very fragmented. Most of society's complex challenges don't get solved with a single data source, so this fragmentation gets in the way of being able to use the world's information to actually make a big difference to everybody everywhere. Google's mission is to organize and make easily accessible all kinds of information. We're doing it now for public data. We've gathered, standardized, and processed data from tens of thousands of tables, from hundreds of reliable global sources, and pulled it into one easily understandable view. This is the fun part. Now comes AI and natural language processing. Most people don't think in terms of scatter plots or variables of interest, or data points or triplets. They think in terms of the question they want to ask of data. Now imagine if you could just simply ask the question in ordinary English, and then you get back these graphs that help give you insights and that let you explore more deeply in sensible ways. That ability to make data accessible and intelligent, pragmatic, practical at the organization level makes Data Commons, I think, an essential tool for organizations as they're doing their work. Generated AI now powers some of the most important functionality in Data Commons. With large language models, you can ask your questions and start getting answers. Imagine, for example, I'm a climate NGO, and I'm trying to understand how access to electricity has changed across countries in Africa. I search for it on Data Commons. From these charts, we see that Kenya, Somalia, and Ghana have had some of the greatest increases in access to electricity. I wonder what the impact of this on their greenhouse gas emissions is going to be. From these charts, we can see that Ghana's greenhouse gas emissions have gone up substantially more than that of Kenya. These observations are the beginning of a journey for us to better understand how we could increase access to things like electricity without proportionally increasing greenhouse gas emissions. When it comes to actually taking action out of the insights from this data, it's going to take people on the ground. We also have to think about this responsibly, so that's why doing this together and thinking about partnerships, collaborations, I think is so so important. It's these people who have deep domain knowledge, who know what the right questions to ask are, who know what the right solutions to build are. At TechSoup, we've been working globally to connect nonprofits to technology for over 30 years. Data is at the heart of what organizations need to be able to make change, so we began working with food banks and food pantries across the United States and Google's Data Commons to build our own instance of Data Commons that pulled together the data points and variables that would be most interesting to these kinds of organizations. One of the big complaints we often hear about small organizations is they don't take risks, and it's hard to take risks if you don't have data because they're operating in the known world, right, the world that they see in their data, which becomes self-reinforcing. Data Commons opens that world up for organizations, means they can take a risk because they're not taking it blindly. I would not have ever guessed I would see this in my lifetime, and it's so exciting to be at a time when this is actually possible. Organizations know the questions they want to ask of data. They have the knowledge that gets at the answers that will be meaningful to them. Data Commons democratizes that. So you just go in and type the question you actually have. That is absolutely game-changing.
|
|
||||||||||