October 17, 2025

news  
Nvidia AI Growth, Salesforce Agentforce, Microsoft Integration

The artificial intelligence landscape continues to evolve rapidly, with new applications and concerns emerging across various sectors. In the gig economy, Uber is piloting a program in the U.S. that allows drivers to earn extra income by completing small AI training tasks, such as recording audio or uploading photos, when they are not driving. This initiative mirrors a similar program already in place in India. Meanwhile, DirecTV is set to offer customers with Gemini devices AI-generated screensavers featuring personalized avatars starting in 2025, complete with interactive shopping links. The broader economic implications of AI are also under scrutiny, with discussions about whether the current surge in AI represents a sustainable boom or a potential bubble, with Nvidia's role in this growth being a key point of analysis. Globally, Europe is working to boost its AI standing, though it trails the U.S. in private investment and computing power, while China leads in open AI models. Addressing the skills gap, IBM and Lynx Educate are partnering to provide free AI and digital skills training worldwide, aiming to train 30 million people by 2030. In the realm of business operations, Pandora is leveraging AI agents to enhance its online sales, improving customer satisfaction and call deflection rates by integrating with systems like Microsoft and Salesforce. HighLevel is also pushing an AI-first vision, showcasing how businesses are scaling with agentic AI, with its platform supporting millions of businesses. However, the advancement of AI also presents challenges. A new investigation reveals the misuse of AI video tools on TikTok to create racist content, raising concerns about content moderation. Furthermore, security vulnerabilities have been identified in Google's Gemini AI, with a flaw that could allow malicious commands, prompting users to disable its access to services like Gmail and Google Calendar. Salesforce has also introduced Agentforce IT Service, utilizing agentic AI for IT support, while Otter.ai has expanded its offerings into an enterprise collaboration suite, integrating with platforms like Microsoft Teams.

Key Takeaways

  • Uber is piloting a program in the U.S. allowing drivers to earn money by completing AI training tasks like uploading photos and audio clips.
  • DirecTV will allow customers with Gemini devices to use AI-generated screensavers featuring their own avatars starting in 2025.
  • Discussions are ongoing regarding whether the current AI growth is a sustainable boom or a potential bubble, with Nvidia's market position being a focus.
  • Europe's AI private investment and computing power lag behind the U.S., though it aims to boost AI adoption.
  • IBM and Lynx Educate are partnering to offer free AI and digital skills training globally, targeting 30 million people by 2030.
  • Racist AI-generated videos are being used on TikTok to gain followers and revenue, highlighting content moderation challenges.
  • A security flaw in Google's Gemini AI could allow malicious commands, with users advised to limit its access to Gmail and Google Calendar.
  • Salesforce has launched Agentforce IT Service, employing agentic AI for automated IT support.
  • Pandora is using AI agents to improve online sales and customer satisfaction, integrating with Microsoft and Salesforce.
  • HighLevel is promoting an AI-first strategy, enabling businesses to scale with agentic AI.

Uber drivers can now train AI for extra cash

Uber is launching a pilot program in the US that lets drivers and couriers earn money by completing small tasks to train AI models. These tasks include recording audio, taking photos, and uploading documents. This initiative is part of Uber's goal to be the top platform for flexible work. The company is also making other app improvements for drivers, such as redesigning trip offer cards and adding a heatmap for high-demand areas. This new feature aims to provide drivers with more ways to earn income.

Uber drivers can earn money training AI

Uber is testing a new program called 'Digital Tasks' in the US, allowing some gig workers to train AI models. This pilot program lets drivers earn extra money by completing tasks like taking photos and uploading audio clips when they are not driving. Uber already offers similar tasks in India. This initiative is part of Uber's broader effort to enhance its platform for flexible work. The company is also improving trip offer details and driver navigation tools.

Uber offers drivers new gig work training AI

Uber is piloting a new program in the U.S. that allows drivers and couriers to earn money through the app when they are not on the road. These tasks involve uploading photos or recording audio to help train AI models, similar to a program already tested in India. Uber CEO Dara Khosrowshahi announced this at the company's Only on Uber 2025 conference. The tasks are separate from autonomous vehicle development and pay varies by complexity. Uber is also introducing other app updates to improve the driver experience.

Is AI a Boom or a Bubble?

This article discusses whether the current excitement around Artificial Intelligence is a sustainable boom or an unsustainable bubble. It references Nvidia and suggests further analysis is needed to understand the long-term economic impact of AI. The piece prompts readers to consider the potential risks and rewards associated with the rapid growth of AI technology.

NPR discusses AI boom and potential bubble

NPR's Ailsa Chang speaks with economist Jared Bernstein about the possibility of an artificial intelligence bubble in the U.S. The discussion explores whether the current rapid growth in AI is sustainable or if it represents an overinflated market. Bernstein, a former chief economic adviser to President Biden, shares his insights on the economic factors at play.

DirecTV customers to see AI versions of themselves in ads

DirecTV is partnering with Glance to let customers using Gemini devices see AI-generated screensavers featuring their own faces starting in 2026. Users can create AI avatars of themselves and even dress them, with shopping links provided for the outfits. These interactive screensavers will offer personalized content and shopping recommendations. The feature activates when the TV is idle for 10 minutes. This move aims to enhance customer interaction with their TV screens through personalized AI experiences.

IBM and Lynx Educate offer free AI training

IBM SkillsBuild and Lynx Educate have partnered to provide free access to AI and digital skills training worldwide. The collaboration aims to train 30 million people by 2030, offering industry-recognized digital badges and project-based learning. This initiative targets individuals without traditional education, frontline workers, and young professionals. The partnership addresses the growing need for updated skills as nearly 40 percent of workers' core skills may need to change by 2030 due to the AI economy.

AI video tools fuel racism on TikTok

TikTok creators are using racist AI-generated videos to gain followers and make money, according to a new investigation. Some videos, depicting violence against migrants, have gained millions of views despite violating TikTok's policies. This raises concerns about content moderation's ability to handle harmful AI-generated content. Several accounts with active TikTok Shops posted these videos, with TikTok removing some content after being contacted. Experts note that advances in AI are being exploited by those with extremist agendas or for financial gain.

Europe's AI standing analyzed in new report

A new report highlights Europe's position in the global AI landscape, showing the US leads significantly in private investment and computing power, while China leads in open AI models. Although the EU has increased its private AI investment, it still lags far behind the US. The report also notes a shift from public to private control of AI supercomputing, potentially limiting academic access. European policymakers are focused on boosting AI adoption and use in science.

Pandora uses AI agents to boost online sales

Pandora, a leading jewelry brand, is using AI agents to enhance its online sales experience, aiming to replicate the personalized service of its physical stores. This new approach has increased customer satisfaction and doubled call deflection rates. The AI agents help customers find jewelry by translating personal interests into relevant product suggestions. Pandora is iteratively cleaning data and integrating AI across its technology stack, including SAP, Microsoft, and Salesforce, to improve the customer journey.

Rowan University lab advances manufacturing with AI

Rowan University's Digital Engineering Hub (DEHub) has launched a new lab focused on integrating artificial intelligence with advanced manufacturing. The lab features 3D printers, scanners, and testing devices, with a key 3D metal printer connected to a supercomputer named Pythia. This setup allows for real-time data processing and flaw correction during manufacturing. The DEHub is initially working on creating a digital twin of the 3D metal printer for the U.S. Department of Defense, with interest from companies worldwide.

HighLevel embraces agentic AI at LevelUp 2025

HighLevel showcased its AI-first vision at its LevelUp 2025 event, highlighting how businesses are scaling with agentic AI. The company's AI Agent Contest saw the creation of hundreds of AI agents for tasks like handling calls and messages across various industries. HighLevel aims to empower AI entrepreneurs by providing tools for businesses to automate and optimize functions. The platform supports over 2 million businesses, processing billions of messages and leads monthly.

Salesforce, Otter, and Gemini AI updates discussed

This tech briefing covers several key updates: Salesforce launched Agentforce IT Service, using agentic AI to automate IT support. Otter.ai expanded its capabilities beyond transcription to become an enterprise collaboration suite, integrating with platforms like Microsoft Teams and Zoom. Experts also warned of a security flaw in Google's Gemini AI that could allow malicious commands through ASCII smuggling attacks, urging users to disable its access to Gmail and Google Calendar until the issue is fixed.

Sources

AI training Gig economy Uber AI models Flexible work AI boom AI bubble Nvidia Economic impact of AI AI avatars Personalized content DirecTV Glance AI skills training IBM SkillsBuild Lynx Educate Digital badges Future of work AI and racism TikTok Harmful content Content moderation AI-generated videos Europe AI US AI investment China AI models AI supercomputing AI adoption AI in manufacturing Rowan University Digital twin 3D printing Agentic AI AI entrepreneurs Business automation Salesforce Agentforce IT Service Otter.ai Enterprise collaboration Gemini AI AI security flaw Google AI