Google's NotebookLM is getting Video Overviews
At Google I/O 2025, the tech giant unveiled new capabilities coming to NotebookLM, its AI-based note-taking and research assistant. Most notably, the company is launching Video Overviews.
The company says users will soon be able to turn dense multimedia, such as raw notes, PDFs, and images, into digestible visual presentations.
Since its launch, NotebookLM has been about helping users understand and interact with complex documents. With this new capability, NotebookLM will be taking a more visual approach to helping users understand different topics and ideas.
NotebookLM has already taken an audio approach to helping users understand materials with Audio Overviews, a feature that gives users the ability to generate a podcast with AI virtual hosts based on documents they have shared with NotebookLM, such as course readings or legal briefs.
Image Credits:Google
Now, Google is rolling out more flexibility to Audio Overviews, as it's letting users select the ideal length for their audio summaries. For example, you can choose to have an Audio Overview at the default length, or longer or shorter.
The new features announced today come a day after Google officially released NotebookLM apps for Android and iOS. Up until now, NotebookLM has only been accessible via desktop. Google has now made the service available on the go.
The apps feature background playback and offline support for Audio Overviews, along with support for dark mode. The apps also allow people to create new notebooks and view the ones they've already created. Plus, when you're viewing a website, PDF, or YouTube video on your device, you can tap the share icon and select NotebookLM to add it as a new source. Users can also view sources that they have already uploaded in each of the notebooks.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
CNET
31 minutes ago
- CNET
Get Your Hands on These Anker Soundcore Open-Ear Earbuds While They're Down to Just $53
There are a lot of options for good running earbuds out there, so finding the one that fits you can be tricky. One really good option is to go with an open-ear style though, as it can help you feel a bit cooler and more comfortable. Well, right now you get your hands on the Soundcore C30i by Anker for $53 as long as you use the on-page coupon to save yourself 25%. This is a great chance to grab these buds for less, and they've got loads of features that make them really good for all kinds of workouts. These open-ear earbuds are available in two colors, black and white. Their firm-shell design is lightweight and comfortable. It also has a secure clip-on design so they won't fall out. You can take them anywhere you go, as they're water-resistant. We haven't tested these specific buds, but they promise clear and high-quality audio, giving you the best listening experience. Hey, did you know? CNET Deals texts are free, easy and save you money. If you're looking for more listening options, check out our list of the best deals on earbuds and headphones going on right now. Why this deal matters Anker's Soundcore open earbuds are designed to fit comfortably in your ears with a tight grip so it doesn't fall out while exercising. The $53 earbuds are now significantly cheaper than Bose's Ultra Open Earbuds which run around $299.
Fast Company
31 minutes ago
- Fast Company
The real data revolution hasn't happened yet
The Gartner Hype Cycle is a valuable framework for understanding where an emerging technology stands on its journey into the mainstream. It helps chart public perception, from the 'Peak of Inflated Expectations' through the 'Trough of Disillusionment,' and eventually up the 'Slope of Enlightenment' toward the 'Plateau of Productivity.' In 2015, Gartner removed big data from the Hype Cycle. Analyst Betsy Burton explained that it was no longer considered an 'emerging technology' and 'has become prevalent in our lives.' She's right. In hindsight, it's remarkable how quickly enterprises recognized the value of their data and learned to use it for their business advantage. Big data moved from novelty to necessity at an impressive pace. Yet in some ways, I disagree with Gartner. Adoption has been widespread, but effectiveness is another matter. Do most enterprises truly have the tools and infrastructure to make the most of the data they hold? I don't believe they do. Which is why I also don't believe the true big data revolution has happened yet. But it's coming. Dissecting the Stack A key reason big data is seen as mature, even mundane, is that people often confuse software progress with overall readiness. The reality is more nuanced. Yes, the software is strong. We have robust platforms for managing, querying, and analyzing massive datasets. Many enterprises have assembled entire software stacks that work well. But that software still needs hardware to run on. And here lies the bottleneck. Most data-intensive workloads still rely on traditional central processing units (CPUs)—the same processors used for general IT tasks. This creates challenges. CPUs are expensive, energy hungry, and not particularly well suited to parallel processing. When a query needs to run across terabytes or even petabytes of data, engineers often divide the work into smaller tasks and process them sequentially. This method is inefficient and time-consuming. It also ends up requiring more total computation than a single large job would. Even though CPUs can run at high clock speeds, they simply don't have enough cores to efficiently handle complex queries at scale. As a result, hardware has limited the potential of big data. But now, that's starting to change with the rise of accelerated computing. Breaking the Bottleneck Accelerated computing refers to running workloads on specialized hardware designed to outperform CPUs. This could mean field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs) built for a specific task. More relevant to big data, though, are graphics processing units (GPUs). GPUs contain thousands of cores and are ideal for tasks that benefit from parallel processing. They can dramatically speed up large-scale data operations. Interestingly, GPU computing and big data emerged around the same time. Nvidia launched CUDA (compute unified device architecture) in 2006, enabling general-purpose computing on graphics hardware. Just two years earlier, Google's MapReduce paper laid the foundation for modern big data processing. Despite this parallel emergence, GPUs haven't become a standard part of enterprise data infrastructure. That's due to a mix of factors. For one, cloud-based access to GPUs was limited until relatively recently. When I started building GPU-accelerated software, SoftLayer—now absorbed into IBM Cloud—was the only real option. There was also a perception problem. Many believed GPU development was too complex and costly to justify, especially for general business needs. And for a long time, few ready-made tools existed to make it easier. Those barriers have largely fallen. Today, a rich ecosystem of software exists to support GPU-accelerated computing. CUDA tools have matured, benefiting from nearly two decades of continuous development. And renting a top-tier GPU, like Nvidia's A100, now costs as little as $1 per hour. With affordable access and a better software stack, we're finally seeing the pieces fall into place. The Real Big Data Revolution What's coming next will be transformative. Until now, most enterprises have been constrained by hardware limits. With GPU acceleration more accessible and a mature ecosystem of supporting tools, those constraints are finally lifting. The impact will vary by organization. But broadly, companies will gain the ability to run complex data operations across massive datasets, without needing to worry about processing time or cost. With faster, cheaper insights, businesses can make better decisions and act more quickly. The value of data will shift from how much is collected to how quickly it can be used. Accelerated computing will also enable experimentation. Freed from concerns about query latency or resource drain, enterprises can explore how their data might power generative AI, smarter applications, or entirely new user experiences. Gartner took big data off the Hype Cycle because it no longer seemed revolutionary. Accelerated computing is about to make it revolutionary again.
Forbes
33 minutes ago
- Forbes
FDA's New AI Tool Cuts Review Time From 3 Days To 6 Minutes
AI at the FDA getty The U.S. Food and Drug Administration announced this week that it deployed a generative AI tool called ELSA (Evidence-based Learning System Assistant), across its organization. After a low-profile pilot that delivered measurable gains, the system is now in use by staff across the agency, several weeks ahead of its original schedule. Dr. Marty Makary, the FDA's commissioner, shared a major outcome. A review task that once took two or three days now takes six minutes. 'Today, we met our goal ahead of schedule and under budget,' said Makary. 'What took one scientific reviewer two to three days [before] The FDA has thousands of reviewers, analysts, and inspectors who deal with massive volumes of unstructured data such as clinical trial documents, safety reports, inspection records. Automating any meaningful portion of that stack creates outsized returns. ELSA helps FDA teams speed up several essential tasks. Staff are already using it to summarize adverse event data for safety assessments, compare drug labels, generate basic code for nonclinical database setup, and identify priority sites for inspections, among other tasks. This last item, using data to rank where inspectors should go, could have a real-world impact on how the FDA oversees the drug and food supply chain and impacts on how the FDA delivers its services. Importantly, however, the tool isn't making autonomous decisions without a human in the loop. The system prepares information so that experts can decide faster. It cuts through the routine, not the judgment. One of the biggest questions about AI systems in the public sector revolves around the use of data and third party AI systems. Makary addressed this directly by saying that 'All information stays within the agency. The AI models are not being trained on data submitted by the industry.' That's a sharp contrast to the AI approaches being taken in the private sector, where many large language models have faced criticism over training on proprietary or user-submitted content. In the enterprise world, this has created mounting demand for "air-gapped" AI solutions that keep data locked inside the company. That makes the FDA's model different from many corporate tools, which often rely on open or external data sources. The agency isn't building a public-facing product. It's building a controlled internal system, one that helps it do its job better. Federal departments have been slow to move past AI experimentation. The Department of Veterans Affairs has started testing predictive tools to manage appointments. The SEC has explored market surveillance AI for years. But few have pushed into full and widespread production. The federal government has thousands of employees processing huge volumes of information, most of it unstructured sitting in documents, files, and even paper. That means AI is being focused most on operational and process-oriented activities. It's shaping up to be a key piece of how agencies process data, make recommendations, and act. Makary put it simply that ELSA is just the beginning for AI adoption within the FDA. 'Today's rollout of ELSA will be the first of many initiatives to come,' he said. 'This is how we'll better serve the American people.'
