Below you’ll see some of the tech trends showcased during NVIDIA GTC that we hear about from TechRepublic.

Advanced generation with advanced search

A technology that aims to reduce AI “hallucinations” or inaccuracies, advanced search generation allows the generative AI model to check its work against external resources, such as scientific papers or documents. RAG appeals to enterprise customers because it increases the reliability of the generated content.

“AI factories” to increase storage and computing needs

Many NVIDIA GTC organizations have described themselves as “AI factories” that provide enterprises with access to the storage and computing power they need to create artificial intelligence.

NexGen Cloud, which calls this service “GPUaaS,” is among the companies that will provide access to NVIDIA’s 10 trillion-parameter Blackwell GPU (Figure A) later this year.

Ten trillion parameter jobs require a lot of compute, and organizations are betting they can create a business model by providing just the right amount of compute power to customers.

During a preview briefing on March 15, Greg Findlen, senior vice president of product management for data management at Dell, shared that data storage must support high-performance structured data as well as unstructured data, such as documents, images and video.

Edge AI

Organizations that focus on edge AI also took their place at NVIDIA GTC 2024, with a wide variety: robotics, automotive, industrial, healthcare, mission-critical systems, and retail.

Many of these were powered by NVIDIA’s Jetson robotics platform. Jetson Orin’s NVIDIA Metropolis microservices allow developers to use API calls to create generative AI capabilities, making robots more reactive and flexible to their environment.

Private AI for enterprises

Organizations are working on creating private generative AI that can securely access their own data while providing the flexibility of a publicly available AI like ChatGPT.

During the show, NVIDIA mentioned Mistral AI several times, which provides a large open source language model that customers can host on their own servers.

Copilots can draw on company-owned data

NVIDIA GTC features a wide range of AI Copilots that can draw answers from specific, company-owned structured and unstructured data.

According to the company, the typical approach to creating a multimodal model involves training components for different formats of information separately and then combining them together. What sets Gemini apart is that it is trained for different formats from the start and then refined with additional multimodal data.

“This helps Gemini seamlessly understand and reason about all kinds of inputs from the ground up, far better than existing multimodal models — and its capabilities are state of the art in nearly every domain,” Sundar Pichai, CEO of Google and Alphabet, and Demis Hassabis, CEO and co-founder of Google DeepMind, wrote in a blog post.

Google also explained that the new model has quite sophisticated reasoning capabilities that allow it to understand complex written and visual information, making it “adept at discovering knowledge that can be difficult to discern among vast amounts of data.”

For example, he can read hundreds of thousands of documents and extract this information to lead to new discoveries in certain fields.

Its multimodal nature makes it particularly suited to understanding and answering questions in complex fields such as mathematics and physics.

Gemini 1.0 offers three distinct versions to cater to various size preferences: Ultra, Pro, and Nano, listed in descending order of size.

According to Google’s initial benchmarking of Gemini, the Ultra version has demonstrated superior performance, outperforming 30 out of the 32 widely used academic benchmarks in model development and research. Notably, Gemini Ultra has achieved a milestone by surpassing human expert performance in massive multitask language understanding (MMLU), encompassing 57 subjects such as math, physics, history, law, medicine, and ethics.

The integration of Gemini Pro into Bard marks a significant milestone, constituting the most substantial update to Bard since its initial launch. The Pixel 8 Pro now leverages Gemini Nano to enhance functionalities like Summarize in the Recorder app and Smart Reply in Google’s keyboard.

Over the coming months, Gemini is set to extend its presence to additional Google products, including Search, Ads, Chrome, and Duet AI.

Starting from December 13, developers can access Gemini Pro through the Gemini API available in Google AI Studio or Google Cloud Vortex AI.

MongoDB Atlas Vector Search is now available, allowing customers to embed generative AI into their applications based on their own data. This enables AI to provide accurate and relevant answers for a specific organization or domain.

Customers can embed artificial intelligence features such as semantic search or image matching into their applications. The new tool uses a “flexible and scalable” document-based data model that the company says allows customers to combine queries for vector data, analytic aggregations, text-based search, geospatial data and time series data.

Alongside this innovative tool, the company has introduced MongoDB Atlas Search Nodes. It offers specialized infrastructure for handling generative AI search tasks associated with MongoDB Atlas Vector Search and MongoDB Atlas Search.

This infrastructure operates independently of the operational nodes in the database, enabling workload isolation, cost efficiency, and improved performance, as elucidated by MongoDB.

“With the general availability of MongoDB Atlas Vector Search and MongoDB Atlas Search Nodes, we’re making it even easier for customers to use a unified, fully managed developer data platform to seamlessly build, deploy, and scale modern applications and provide end users with the types of personalized, AI-powered experiences that save them time and keep them engaged,” said Sahir Azam, chief product officer at MongoDB.

Illustratively, a retailer conducting a holiday promotion could leverage this infrastructure to isolate and scale workloads specifically for chatbots in a designated location.

MongoDB asserts that this novel service has the potential to decrease query times by as much as 60%.

This feature enables the development of headless content applications that can integrate various management functionalities, such as collaboration, security and compliance, into each application.

Enterprises can use SharePoint Embedded to create business applications, offering a unified experience for both users and the system administrators who manage those applications.

Similarly, ISVs can use this technology to incorporate Microsoft 365 content management capabilities into any enterprise application they develop.

With SharePoint Embedded, documents are managed within the customer’s Microsoft 365 tenant, providing a consistent content management system with global security and compliance features.

SharePoint Embedded serves as a versatile platform intended for utilization by both Microsoft’s in-house products and third-party applications developed by any developer. This information is detailed in a blog post by the company, which outlines the comprehensive list of features in SharePoint Embedded.

Notably, Microsoft Loop and Microsoft Designer, two recently introduced apps, harness the capabilities of SharePoint Embedded. The platform is constructed on the same stack as SharePoint and OneDrive, distinguishing itself through its remarkable scalability, performance, and aptness for hyperscale use cases.

Integration of sophisticated Microsoft 365 security and compliance features, such as audit, eDiscovery, Bring Your Own Key, and others, is facilitated for developers. Additionally, the platform allows for the utilization of Microsoft 365 business continuity capabilities, encompassing autoscaling, high availability, Microsoft 365 Backup, Microsoft 365 Archive, and disaster recovery.

Furthermore, developers have the capability to integrate Microsoft 365 collaboration features into their applications, including co authoring in Word or PowerPoint. They can also embed trusted content management functionalities such as search, content preview, and version tracking. Additionally, the platform allows for the customization of user experiences, providing complete control over security, lifecycle, and management aspects.

If this applies to you and you want to keep your account, continue reading.

The tech giant says it will delete all content from the accounts in question, from emails to photos and various documents created and stored on the platform. So, if you’ve left an old account that you don’t use, but don’t want its content to disappear forever, you can save your account and keep it intact. Want to find out?

How to save your Google account?
To “reactivate” an account, simply log in to it once, without taking any particular action. Simply logging in and performing a Google search, reading an email, looking at YouTube or Drive, or downloading a mobile application using this account is enough to make it active again. In all cases, Google will send several notifications before deleting an account, including to a recovery address, if one has been provided.

Note that this measure only applies to personal accounts, not those linked to an employer, school or other organization. Nor does it apply to anyone with a paid-for Google One account.

To avoid any unpleasant surprises in the weeks to come, we strongly advise you to log in to your old, sometimes forgotten, accounts. And if you find any important documents there, you can transfer them to your main account by going to Google Takeout. Select all the data you want to download, then copy it service by service to your other account. Then you won’t have to worry about keeping your old account active.

“We are excited to announce AI integration in multiple products to enhance developer productivity and creativity. At the same time, we’ve added one of the most-often requested features to MapForce, which is support for PDF. The new MapForce PDF Extractor will be a game changer for unlocking the volumes of data previously trapped in PDF documents and accessing it for use in data integration and ETL processes”, said Alexander Falk, president and CEO of Altova.

Altova announced the latest additions to its product line, featuring cutting-edge AI capabilities that promise to redefine the landscape of XML, JSON development, and database management. Among these innovations, the spotlight shines on the AI Assistant integrated into XMLSpy, a feature designed to significantly boost productivity.

The AI Assistant in XMLSpy streamlining XML and JSON development by effortlessly generating schemas, instance documents, and sample data through intuitive natural language prompts.

XMLSpy can also produce XSL, XPath, and XQuery code. Users have the flexibility to copy generated code, open it in a new document, or send it to the XPath/XQuery window for further tweaking and examination.

In DatabaseSpy, Altova introduces AI integration that augments the database management experience. The AI Assistant within DatabaseSpy equips users with the ability to effortlessly generate SQL statements, sample data, and table relations. Additionally, AI extensions empower users to explain, format, and complete SQL statements with unparalleled ease and efficiency.

Further enhancements include the MapForce PDF Extractor, a visual utility that simplifies the definition of a PDF document’s structure and streamlines data extraction. In StyleVision, the new Split output preview feature for XML and database report design provides users with a comprehensive view of their output in side-by-side panels, further enhancing the development process.

ESET’s research found that the Android app “iRecorder – Screen Recorder” introduced the malicious code as an app update almost a year after it was first listed on Google Play. According to ESET, the code allowed the app to invisibly upload one minute of ambient audio from the device’s microphone every 15 minutes, as well as extract documents, web pages and media files from the user’s phone.

Google Play has removed the app from its list. It’s a good idea to delete the app if it was installed on your device. The malicious app has accumulated more than 50,000 downloads.

ESET is calling the malicious code AhRat, a customized version of an open-source remote access trojan called AhMyth. Remote access trojans (or RATs) take advantage of broad access to a victim’s device and can often include remote control, but also function similarly to spyware and stalkerware.

In fact, AhMyth isn’t sneaking onto Google Play for the first time. Both Google and Apple check apps for malware before listing them for download. Last year, Google said it prevented more than 1.4 million privacy-violating apps from accessing Google Play.

The FTC is cracking down on GoodRx, Microsoft is dropping “verified” phishing scams, researchers are uncovering vulnerabilities in electric vehicle chargers, and more, Wired wrote on the topic.

Spam and malware tracking NGO Spamhaus reports that it has seen a “huge spike” in the spread of malware via Google ads over the past two months. This includes “malware” that looks like authentic downloads of tools such as Slack, Mozilla’s Thunderbird email client and the Tor browser.

Security firm SentinelOne further identified a handful of malicious bootloaders distributed through Google ads, which the researchers collectively dubbed MalVirt. They say the MalVirt loaders are used to spread malware such as XLoader, which an attacker can use to steal data from an infected machine.

This week, Microsoft announced that it has disabled the accounts of threat actors who were able to verify themselves under the Microsoft Cloud Partner Program. Posing as legitimate companies, threat actors used their verified account status to create malicious OAuth applications.

The 37 million customers put at risk by the latest T-Mobile hack may not be the only people affected by the breach. This week, Google informed customers of its Google Fi mobile service that hackers had obtained “limited” account information, including phone numbers, SIM card serial numbers and information about their accounts.

The hackers did not gain access to payment information, passwords or the content of messages, such as text messages. However, the information may have been used for SIM replacement attacks.

Couchbase Capella is a fully managed database with JSON documents and key-values with SQL access and built-in full-text search, event creation and analytics. It can be distributed globally via AWS and GCP through their data centers and availability zones. In addition, the company now provides the database in Azure, allowing its customers to flexibly use Capella across all three major cloud service providers, Infoq wrote on the topic.

The company made the service available first on AWS and has since made several enhancements to the DBaaS offering, including a new developer user experience, SOC 2 Type 2 compatibility, a fully managed backend designed for mobile, IoT and edge applications with App Services, and a new Magma HD storage engine that gives customers the ability to support more use cases with much larger data sets at high speed.

Microsoft offers Azure Cosmos DB, a competitive service comparable to Couchbase Capella. Couchbase previously compared the two offerings, with several differences, such as that Capella is memory-based, while CosmosDB is caching-based, and consumption is on a per-node basis versus per-unit queries.

Finally, developers can try out Capella by signing up for a 30-day period. Finally, customers who want to use Capella in Azure can purchase the offering from the Azure marketplace.

Machine learning has a valuable role to play in legal disputes. Legal technology is increasing efficiency, reducing delays and lowering costs by making legal services more accessible than ever. And whether or not it’s approved by most people, artificial intelligence is gradually starting to play an important role in the legal system and dispensing justice.

Lawyer365 (UK)
Lawyer365 brings teleconsulting to law with a digital app providing access to free legal information services and free appointments with qualified lawyers, straight from your smartphone. Lawyer 365 lets you see a lawyer by video 365 days a year in minutes.

Legal document templates are available, including hundreds of editable agreements and contracts to enable you to protect your business and keep up with your legal and compliance requirements.

The company has raised $1.8M in funding over two rounds, including a seed round in April 2022.

Avokaado (Estonia)
If there’s one workplace task begging for legal automation, it’s the drafting of legal documentation. Avokaado has developed automated templates that can draft even the most complex documents and contracts up to 98% faster.

The company raised €700,000 in pre-seed funding in 2021.

Chi Odia Paga (Italy)
Cybercrimes happen every day, but the laws in many countries are decades behind rapidly evolving technology. In Italy, conduct such as stalking, revenge porn, defamation, and threats, if prosecuted, can result in fines and even prison sentences.

Chi Odia Paga offers a quick and functional digital service to fight hate online and help people who have suffered legally prosecutable conduct obtain justice and adequate compensation, all without upfront payments.

Eperoto (Sweden)
Legal disputes can have enormous financial and emotional challenges for all parties involved. Swedish startup Eperoto was founded in 2020 and has developed an easy-to-use software tool to help those in legal disputes to attain a structured analysis of the potential courses of action. This enables users to make rational and informed decisions whether in a local court or international arbitration case.

The company has raised $351.6K in funding over three rounds, with a pre-seed round in 2021.

DeepJudge (Switzerland)
Law firms typically work on hundreds of documents per case, which are nothing short of unwieldy and time-consuming. DeepJudge AG is a spin-off of ETH Zurich focusing on the contextual understanding and processing of legal documents.

Its technology combines Deep Learning and Natural Language Processing to create more comprehensive documents. For example, it automatically highlights crucial information in documents, referencing with relevant external data such as legal codes, court rulings, commercial registries, encyclopedias and etc.

Uncover (The Netherlands)
Dutch startup Uncover has developed software to help litigation lawyers through every step of court preparation. Leveraging AI, Uncover software can readily process large amounts of documentation and, for example, accurately identify relevant events to generate a timeline, which is often a foundation for building a solid legal case.

Conny (Germany)
Berlin-based Conny was founded in 2016 and has built a tech platform that makes consumer legal claims. Whether you’re fighting a rent increase, bond return, or unlawful termination at work, Conny strives to make the process as easy and fast as ordering a pizza. Its software checks possible claims and the chances of success free of charge in the shortest possible time, and it usually only costs you if you win your case.

The company raised an undisclosed Series A in 2019.