Read AI Archives - gettectonic.com
Apple's Privacy Changes: A Call for Email Marketing Innovation

Liar Liar Apple on Fire

Apple Developing Update After AI System Generates Inaccurate News Summaries Apple is working on a software update to address inaccuracies generated by its Apple Intelligence system after multiple instances of false news summaries were reported. The BBC first alerted Apple in mid-December to significant errors in the system, including a fabricated summary that falsely attributed a statement to BBC News. The summary suggested Luigi Mangione, accused of killing United Healthcare CEO Brian Thompson, had shot himself, a claim entirely unsubstantiated. Other publishers, such as ProPublica, also raised concerns about Apple Intelligence producing misleading summaries. While Apple did not respond immediately to the BBC’s December report, it issued a statement after pressure mounted from groups like the National Union of Journalists and Reporters Without Borders, both of which called for the removal of Apple Intelligence. Apple assured stakeholders it is working to refine the technology. A Widespread AI Issue: Hallucinations Apple joins the ranks of other AI vendors struggling with generative AI hallucinations—instances where AI produces false or misleading information. In October 2024, Perplexity AI faced a lawsuit from Dow Jones & Co. and the New York Post over fabricated news content attributed to their publications. Similarly, Google had to improve its AI summaries after providing users with inaccurate information. On January 16, Apple temporarily disabled AI-generated summaries for news apps on iPhone, iPad, and Mac devices. The Core Problem: AI Hallucination Chirag Shah, a professor of Information Science at the University of Washington, emphasized that hallucination is inherent to the way large language models (LLMs) function. “The nature of AI models is to generate, synthesize, and summarize, which makes them prone to mistakes,” Shah explained. “This isn’t something you can debug easily—it’s intrinsic to how LLMs operate.” While Apple plans to introduce an update that clearly labels summaries as AI-generated, Shah believes this measure falls short. “Most people don’t understand how these headlines or summaries are created. The responsible approach is to pause the technology until it’s better understood and mitigation strategies are in place,” he said. Legal and Brand Implications for Apple The hallucinated summaries pose significant reputational and legal risks for Apple, according to Michael Bennett, an AI adviser at Northeastern University. Before launching Apple Intelligence, the company was perceived as lagging in the AI race. The release of this system was intended to position Apple as a leader. Instead, the inaccuracies have damaged its credibility. “This type of hallucinated summarization is both an embarrassment and a serious legal liability,” Bennett said. “These errors could form the basis for defamation claims, as Apple Intelligence misattributes false information to reputable news sources.” Bennett criticized Apple’s seemingly minimal response. “It’s surprising how casual Apple’s reaction has been. This is a major issue for their brand and could expose them to significant legal consequences,” he added. Opportunity for Publishers The incident highlights the need for publishers to protect their interests when partnering with AI vendors like Apple and Google. Publishers should demand stronger safeguards to prevent false attributions and negotiate new contractual clauses to minimize brand risk. “This is an opportunity for publishers to lead the charge, pushing AI companies to refine their models or stop attributing false summaries to news sources,” Bennett said. He suggested legal action as a potential recourse if vendors fail to address these issues. Potential Regulatory Action The Federal Trade Commission (FTC) may also scrutinize the issue, as consumers paying for products like iPhones with AI capabilities could argue they are not receiving the promised service. However, Bennett believes Apple will likely act to resolve the problem before regulatory involvement becomes necessary. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Service Cloud with AI-Driven Intelligence Salesforce Enhances Service Cloud with AI-Driven Intelligence Engine Data science and analytics are rapidly becoming standard features in enterprise applications, Read more

Read More
Agentforce to the Team

Agentforce to the Team

Salesforce has introduced the Agentforce Atlas Reasoning Engine, a platform designed to perform tasks autonomously with minimal human intervention. Agentforce to the Team changes everything about AI. Businesses can feed the engine data, assign tasks, and step away, as the system is capable of completing work independently. This launch closely follows OpenAI’s recent advancements in artificial intelligence, highlighting the ongoing collaboration between Salesforce and Sam Altman’s firm. Agentforce to the Team-makes me hear “Honey, I’m home”, coming from the front door. The Agentforce Atlas Reasoning Engine is designed to analyze data, make decisions, and execute tasks with high reliability and accuracy, echoing the features of OpenAI’s latest AI model. Salesforce positions this as part of the “Third Wave of AI,” where intelligent agents go beyond assisting humans to actively driving business outcomes without frequent oversight. According to Salesforce CEO Marc Benioff, these agents are deeply integrated into customer workflows, anticipating needs and improving growth by taking proactive action at every touchpoint. Benioff emphasized the revolutionary nature of Agentforce, which he claims will surpass existing AI platforms by offering highly accurate, low-hallucination results. It integrates seamlessly across Salesforce’s ecosystem, benefiting users from industries such as financial services, healthcare, and government. Early adopters, such as Wiley, report a 40% increase in case resolution, with Agentforce handling routine customer service tasks more efficiently than previous chatbots. Disney also saw improved results, noting that Atlas delivered twice the accuracy of other AI tools they had benchmarked. However, the autonomous nature of these agents raises concerns about job displacement, particularly for workers involved in repetitive, low-impact tasks. While Salesforce advocates for reskilling workers to transition into higher-value roles, many organizations struggle to effectively implement such initiatives. The time required to upskill workers may not align with the rapid adoption of AI technologies like Agentforce. Agentforce aims to address common enterprise challenges by offering out-of-the-box solutions for sales, marketing, and customer service roles. The low-code platform allows businesses to customize their AI agents without extensive technical expertise, ensuring that they can scale capacity and improve efficiency. Salesforce plans to showcase Agentforce at its upcoming Dreamforce conference, aiming to onboard 1,000 customers to the platform. The launch signifies Salesforce’s strategic push to dominate the enterprise AI landscape, leveraging its vast data and platform to deliver more value to its customers. Despite its potential, Agentforce introduces new risks, especially in areas like data privacy and ethical AI deployment. Salesforce emphasizes its commitment to addressing these issues by incorporating ethical guardrails, such as toxicity filters. Industry analysts remain cautiously optimistic, noting that while the technology holds promise, the real test will come as more organizations adopt it and integrate it into their workflows. In summary, Salesforce’s Agentforce Atlas Reasoning Engine represents a significant leap in enterprise AI, moving beyond basic AI copilots to fully autonomous agents. While it offers substantial benefits in productivity and efficiency, its impact on the workforce and the challenges of widespread AI adoption will require ongoing attention. By Tectonic’s Shannan Hearne, Solutions Architect Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Service Cloud with AI-Driven Intelligence Salesforce Enhances Service Cloud with AI-Driven Intelligence Engine Data science and analytics are rapidly becoming standard features in enterprise applications, Read more

Read More
Read AI Salesforce Integration

Read AI Salesforce Integration

Last month, Read AI announced the launch of its integration with HubSpot, and today the company continues its momentum in the CRM space by announcing support for 1-click connections to Salesforce. Salesforce, the leading name in CRM software, is widely adopted across virtually every industry. Read AI users can now sync their Meeting Reports directly to corresponding records within their Salesforce instance, either automatically or manually. After a meeting measured by Read AI, this integration will automatically connect any Contacts, Leads, Accounts, and Opportunities within Salesforce that are related to the external participants who attended the call. Read AI will sync Meeting Report summaries, action items, key questions, and other meeting data directly to those records, providing a comprehensive view of meeting content and progress without leaving Salesforce. This eliminates the need to spend hours updating opportunities. With Read AI, Opportunities are more up-to-date and comprehensive than anything manually written. The integration operates in the background, automatically. For those who prefer not to automatically sync Meeting Reports into Salesforce, Automatic Syncing can be turned off in Integration Settings, allowing individual push of Meeting Reports to Salesforce. To ensure historical Meeting Reports are synced to Salesforce, users can backfill Meeting Report data with one click using the ‘Sync past meetings’ button in Salesforce Integration Settings. For individual sellers, this integration removes the need to block out time to update contacts and opportunities, as Read AI automatically makes those updates. Sales Managers will no longer need to chase down sellers to update their opportunities in Salesforce, allowing more focus on sales strategy with the most complete and timely insights from SFDC. This integration promises a significant boost in productivity for users of both Read AI and Salesforce. Users no longer need to take notes during important sales calls and can easily find meeting details within Salesforce. Read AI empowers sellers to close deals while handling the administrative tasks to ensure opportunities are accurate and updated. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Service Cloud with AI-Driven Intelligence Salesforce Enhances Service Cloud with AI-Driven Intelligence Engine Data science and analytics are rapidly becoming standard features in enterprise applications, Read more

Read More
gettectonic.com