The Comunicano for Thursday March 27th 2025
Thirsty Thursday
We live in an age of astonishing convenience. A swipe here, a tap there — and the world responds. We plan trips, order groceries, even see doctors from the glow of our screens. But somewhere between the scroll and the like, a quiet question begins to rise: who’s really holding the reins?
Take TikTok. Just a year ago, half the country wanted it gone — a foreign-owned app soaking up our lives, our data, our children’s attention. Now? Support for a ban is slipping. Maybe we’ve grown used to it. Maybe we’ve just grown tired. But the ownership hasn’t changed. The risks haven’t changed. Only our attention span has.
Meanwhile, halfway around the world, companies with hidden hands are trying to hire Americans who once served their country. Not with offers of pride or purpose, but with paychecks. It’s clever. It’s quiet. It’s something to think about.
And as we dive deeper into the digital future — cloud breaches, photorealistic AI art, robots the size of rice grains — the line between science fiction and Monday morning is fading fast. We see the beauty. The promise. But we must also see the responsibility.
Technology should serve us — not replace us, not surveil us, not sell us out.
There’s no going back to simpler times. But we can go forward with our eyes open. We can demand transparency, protect our people, and teach our kids not just how to code — but how to care about the country they’re coding in.
Because in the end, this isn’t about apps, or models, or machines.
It’s about us. Our values. Our future.
And whether we’re ready to hold on to both. That’s why you read The Comunicano!!!
Andy Abramson
TikTok Watch
Public Support for TikTok Ban Declines, National Security Concerns Ease (Pew Research Center)—A recent Pew Research Center survey indicates that 34 percent of U.S. adults support a government ban on TikTok, down from 50 percent in March 2023. The perception of TikTok as a national security threat has also decreased from 59 percent to 49 percent over the same period. Among supporters of a ban, data security and the platform's Chinese ownership are primary concerns, while opponents often cite free speech implications. These findings come ahead of an April deadline for TikTok to be sold or face a nationwide ban. Read more here
Fed Watch
Secretive Chinese Network Attempts to Recruit Former U.S. Federal Employees (The Japan Times)—A clandestine network of companies, operated by a Chinese technology firm, has been attempting to recruit recently laid-off U.S. government workers. Max Lesser, a senior analyst at the Foundation for Defense of Democracies, uncovered that these entities, posing as consulting and headhunting firms, are targeting former federal employees and AI researchers through deceptive job advertisements. The network's methods mirror known Chinese intelligence operations that exploit financial vulnerabilities of unemployed individuals. Despite investigations, direct ties to the Chinese government remain unconfirmed, and it is unclear if any recruitment efforts have been successful. Read more here
Security Watch
Oracle Customers Validate Data Stolen in Alleged Cloud Breach (BleepingComputer)—Despite Oracle's denial of a breach in its Cloud federated SSO login servers, BleepingComputer confirmed with multiple companies that data samples shared by the threat actor are authentic. The hacker, 'rose87168,' claims to have accessed information on 6 million users, including encrypted passwords and a list of 140,621 domains associated with various companies and government agencies. While Oracle maintains that no breach occurred, affected companies have verified the legitimacy of the leaked information, raising concerns about the security of Oracle's cloud infrastructure. Read more here
OpenAI Watch
OpenAI Introduces 4o Image Generation, Enhancing ChatGPT's Visual Capabilities (OpenAI)—OpenAI has unveiled 4o Image Generation, a feature integrated into the GPT-4o model, enabling ChatGPT to produce precise, photorealistic images. This advancement allows for the creation and editing of images directly within ChatGPT, offering users a seamless multimodal experience. The model's deep integration facilitates nuanced and expressive image generation, expanding the utility of AI in creative and professional contexts. Access to this feature is rolling out to Plus, Pro, Team, and Free users, with Enterprise and Edu access forthcoming. Read more here
ChatGPT's New Image Feature Sparks Trend of Studio Ghibli-Style Portraits (Axios)—Following the release of ChatGPT's image-generation capabilities powered by GPT-4o, users have begun creating portraits in the distinctive style of Studio Ghibli. This trend has gained momentum on social media platforms, with individuals transforming personal photos into Ghibli-inspired artwork. The feature's ability to emulate specific artistic styles has elicited mixed reactions, with some users praising the creativity and others criticizing the quality of AI-generated art. OpenAI's relaxation of previous restrictions on generating images in distinctive styles has contributed to this surge in Ghibli-style creations. Read more here
Microsoft Watch
Microsoft Enhances 365 Copilot with 'Deep Reasoning' AI Agents (The Verge)—Microsoft has introduced two advanced AI agents, Researcher and Analyst, to its Microsoft 365 Copilot suite. Researcher utilizes OpenAI's deep research model to perform complex, multi-step research tasks and integrates data from third-party sources like Salesforce and ServiceNow. Analyst, based on OpenAI's o3-mini reasoning model, can transform raw data into spreadsheets, execute Python code, and generate comprehensive reports, functioning similarly to a skilled data scientist. These features will be available to Microsoft 365 Copilot license holders starting in April through an early access program. Additionally, new autonomous agent capabilities are being introduced in Copilot Studio, enabling the automation of various tasks through AI-driven workflows. Read more here
AI Watch
AI Crawlers Overwhelm Web Traffic, Prompting Developers to Block Entire Countries (Ars Technica)—Open-source developers report that AI data crawlers are generating excessive web traffic, leading to increased server costs and operational disruptions. The 'Read the Docs' project experienced a 75% traffic reduction after blocking AI crawlers, highlighting the extent of the issue. Some developers have resorted to blocking entire countries to mitigate the impact, reflecting the challenges posed by uncontrolled AI data scraping on web infrastructure and the need for more responsible crawling practices by AI companies. Read more here
China's AI Data Centers Remain Underutilized Amid Shifting Industry Trends (MIT Technology Review)—China's rapid construction of AI-focused data centers has led to an oversupply, with reports indicating that up to 80 percent of newly built computing resources are unused. Initially driven by government and private investment to capitalize on the AI boom, the industry now faces challenges due to speculative investments, weak demand, and shifts in AI development trends, such as the rise of models like DeepSeek. This situation has resulted in financial struggles for many data center operators, leading to potential government intervention to manage these underutilized assets. Read more here
As AI Use Grows at Work, On-the-Job Training Lags (Yahoo Finance)—The adoption of artificial intelligence in the workplace has surged, with 35 percent of workers now utilizing AI tools, up from just 8 percent two years ago. Despite this rapid increase, over half of these employees report feeling unprepared to effectively integrate AI into their roles. This gap highlights the urgent need for comprehensive on-the-job training programs to equip workers with the necessary skills to harness AI technologies, ensuring that the workforce remains competent and competitive in an evolving technological landscape. Read more here
Best AI Email Assistants Ranked: Microsoft Edges Out Google (Washington Post)—In a comparative test of AI-powered email tools, the Washington Post evaluated offerings from Microsoft, Google, Apple, and others to see which assistant actually saves time. Microsoft 365 Copilot was praised for its ability to write drafts in a user’s voice and summarize long threads efficiently. Google’s Gemini, while capable, lagged in real-world usefulness. Apple’s AI, still in the early stages, failed to deliver substantive help. As AI email tools proliferate, their performance and user-friendliness are shaping their impact in daily professional life. Read more here
Bill Gates Predicts AI Could Replace Humans in Most Jobs Within a Decade (CNBC)—Bill Gates has suggested that within the next 10 years, artificial intelligence could advance to a point where humans are no longer needed for most tasks. He emphasized that AI's rapid development may lead to significant changes in the workforce, potentially rendering many human roles obsolete. Gates highlighted the importance of preparing for this shift by focusing on education and policy adjustments to address the challenges posed by widespread automation and AI integration in various industries. Read more here
Global Shift Away from U.S. AI Models Amid Privacy and Security Concerns (MIT Technology Review)—Countries worldwide are increasingly seeking alternatives to U.S.-developed AI models, driven by concerns over data privacy, security, and the dominance of American tech companies. This movement, termed "AI nationalism," reflects a desire for technological sovereignty and the protection of national interests. Nations are investing in domestic AI capabilities and implementing regulations to reduce reliance on foreign technologies. This trend has significant implications for the global AI landscape, potentially leading to a more fragmented market and influencing international collaborations and competition in artificial intelligence development. Read more here
Google Watch
I Quit Google Search for AI—and I'm Not Going Back (The Wall Street Journal)—In a personal experiment, a journalist replaced Google Search with AI chatbots like ChatGPT, Perplexity, and Google's Gemini for daily queries. The AI tools provided concise answers and useful links, reducing exposure to ads and clickbait common in traditional search results. While AI chatbots enhanced efficiency in tasks like shopping and recipe searches, concerns remain about misinformation and the impact on content creators' visibility. Despite these issues, the experience suggests a significant shift in how users may seek information online. Read more here
Google Unveils Gemini 2.5, Its Most Advanced AI Model Yet (Computerworld)—Google has introduced Gemini 2.5, an AI model that surpasses previous iterations in reasoning and multimodal capabilities. Gemini 2.5 can analyze information, draw logical conclusions, and make informed decisions by interpreting text, audio, images, video, and code. This advancement positions Google ahead of competitors like OpenAI and Anthropic, offering developers enhanced tools for creating applications and games. The model's improved reasoning abilities enable more accurate and context-aware responses, marking a significant step forward in artificial intelligence development. Read more here
Google Adds Vacation Planning Tools to Search, Maps, and Gemini (TechCrunch)
Google is rolling out new vacation planning features across Search, Maps, and Gemini, aiming to streamline travel research and trip organization. In Search, users can now browse curated vacation ideas with visuals, itineraries, and activity suggestions. Maps gains enhancements to the "Explore" tab, highlighting top sights, dining, and personalized recommendations based on travel dates. Gemini, Google’s AI assistant, can now assist with creating day-by-day itineraries, suggesting hotels, activities, and logistics in a conversational format. These updates aim to make planning travel as intuitive and integrated as possible within Google's ecosystem. Read more here
Waze Removes Google Assistant from iOS Due to Persistent Issues (TechCrunch)—Waze has officially discontinued Google Assistant integration on its iOS app, citing unresolved technical difficulties that have persisted for over a year. Users reported issues where the voice command feature failed to respond appropriately. In response, Waze plans to introduce an "enhanced voice interaction solution" in the near future. This decision aligns with Google's broader strategy to transition users from Google Assistant to its new AI model, Gemini, which is expected to replace Assistant on most mobile devices later this year. Read more here
Amazon Watch
Amazon Tests AI-Powered Shopping and Health Assistants (CNBC)—Amazon is piloting two AI-driven tools: "Interests," a personalized shopping assistant, and "Health AI," a health-focused chatbot. "Interests" allows users to input prompts in everyday language to receive tailored product recommendations, enhancing the shopping experience. "Health AI" provides health information and guidance, helping customers find products and solutions for their healthcare needs. These initiatives reflect Amazon's ongoing investment in generative AI to improve customer engagement and service offerings. Read more here
Retail Watch
Instacart Introduces 'Store View' and 'Second Store Check' to Enhance Inventory Accuracy (The Verge)—Instacart is launching two new features aimed at improving inventory accuracy and customer satisfaction. 'Store View' allows shoppers to earn money by capturing videos of store shelves, providing real-time data on product availability and out-of-stock items. This initiative helps Instacart predict restocking times more accurately. Additionally, 'Second Store Check' enables a secondary shopper to source out-of-stock items from nearby locations, ensuring customers receive their desired products without additional fees. These innovations are part of Instacart's broader effort to leverage technology for a more reliable grocery delivery experience. Read more here
Robot Watch
China's Position in the Global Robotics Race Faces Challenges (Reuters)—China is emphasizing robotics to address its aging workforce and labor shortages. Despite leading in industrial robot installations, many are low-tech compared to advanced humanoid robots under development globally. Companies like Tesla and Nvidia anticipate widespread use of sophisticated robots soon. However, China's reliance on foreign components and potential U.S. sanctions pose risks to its robotics ambitions. While progress is evident, China's dominance in the robotics sector remains uncertain due to these dependencies. Read more here
Innovative Liquid Robots Mimic Biological Cells (Singularity Hub)—Researchers in Korea have developed rice-sized liquid robots made of water coated with Teflon particles, inspired by biological cells. Controlled using sound waves, these robots can split, merge, transport cargo, and function as tiny chemical reactors. Their biocompatibility suggests potential applications in targeted drug delivery and diagnostics within the human body. This advancement represents a significant step in creating adaptable, soft robots for medical and technological uses. Read more here
Media Watch
Boston-Area Newsrooms Implement AI-Assisted Reporting, Sparking Debate (Boston.com)—Several Gannett-owned news outlets in the Boston area, including The Patriot Ledger and the Cape Cod Times, have begun utilizing Espresso, a generative AI tool designed to assist reporters in drafting articles from community announcements. This initiative has led to the emergence of roles such as "AI-assisted reporter," exemplified by Beth McDermott, who has authored numerous articles across these publications. The adoption of AI in journalism has prompted discussions about the future of reporting and the role of technology in fostering community engagement. Read more here
FinTech Watch
Robinhood Launches Online Banking with On-Demand Cash Delivery (The Verge)—Robinhood is expanding its financial services by introducing Robinhood Banking, an online platform offering checking and savings accounts to Robinhood Gold subscribers starting this fall. The service features a 4 percent annual percentage yield and FDIC insurance up to 2.5 million dollars through Coastal Community Bank. A standout feature is the on-demand delivery of physical cash to users' doorsteps, though details on implementation remain limited. This initiative reflects Robinhood's strategy to provide comprehensive financial solutions within its ecosystem. Read more here
