Google introduces Gemini agent that can serve as your AI coworker: Here is what it can do

digit.in Oct 9, 2026

Google introduces Gemini agent that can serve as your AI coworker: Here is what it can do

Curated by Shubham · TechCapsules · Oct 9, 2026

AI-summarised brief · reviewed before publication

Google unveiled its Gemini agent at the Gemini at Work 2026 event, positioning it as an AI coworker capable of answering questions, generating images, writing and executing code, and integrating with existing workflows. The agent can create specialized sub‑agents with distinct roles and custom email addresses, and it employs four memory types to retain context. Users can select from multiple models to balance performance and cost while maintaining enterprise security. Gemini is available across Android, iOS, web, Windows, macOS, and platforms like Microsoft 365, Slack, and command‑line interfaces, and can be embedded into third‑party applications. Google also announced industry‑specific Gemini versions for legal and financial services, with more sectors to follow.

💡 Why It Matters

· By turning AI into a modular coworker that can be tailored to specific roles and integrated into everyday tools, Google is redefining workplace automation and potentially accelerating productivity gains across industries.

AI Capsules Google

Read full article on digit.in

Ben Affleck’s AI expertise goes viral, reveals how tech could transform filmmaking

Ben Affleck, known for his Batman role, has sparked viral interest by detailing his knowledge of AI technologies that could reshape filmmaking. In interviews with GQ and Bloomberg, he explained neural networks, Python scripting, and transformer models, emphasizing how these tools can aid visual effects rather than replace artists. Affleck highlighted practical applications such as background removal and object isolation, and stressed the need for safeguards and creative control. His comments arrive amid Hollywood debates over generative AI’s impact on creative jobs, likenesses, and production economics.

💡 Affleck’s technical insights bridge a gap between celebrity influence and emerging AI tools, offering a pragmatic perspective that could guide industry standards for responsible adoption.

risingnepaldaily.com · Oct 9

AI Weakens Minds

Artificial intelligence has surged from a niche tool to a pervasive force shaping media, advertising, and social platforms. Recent reports indicate that roughly 20 % of new YouTube videos are AI‑generated, while AI‑created thumbnails and full-length content now dominate many channels. The trend extends to Instagram and TikTok, where fabricated clips and low‑effort AI “slop” videos attract millions of views. This shift erodes traditional creative labor, encourages plagiarism, and spreads misleading content, raising concerns about authenticity and intellectual integrity in digital media.

💡 The rise of AI‑slop threatens to dilute genuine creativity, eroding trust in online content and undermining the value of human effort in media production.

kutv.com · Oct 9

As AI races forward, Utah researchers work to keep humans in control

Artificial intelligence is advancing rapidly, prompting former employees of leading AI firms to warn that safety safeguards are insufficient. At the University of Utah, researchers are exploring human‑AI collaboration to ensure control remains with people. In a robotics lab, Ph.D. student Connor Mattson programs low‑cost robots to learn everyday tasks such as cooking, emphasizing assistance rather than replacement. Simultaneously, Ph.D. candidate Zohre Karimi trains a larger robotic arm to excise throat tumors, arguing that surgeons could treat more patients while staying in command of the procedure. Assistant professor Daniel Brown notes the university’s participation in the Center for Human and Robot Co‑Adaption, a consortium of six schools investigating safe, cooperative robot design as companies race to build powerful AI future systems.

💡 Keeping humans in the decision loop is essential to prevent autonomous systems from making unchecked choices as AI becomes integral to healthcare and household tasks.

computerweekly.com · Oct 9

AI model developers have ‘no justification’ for failing to comply with privacy…

UK data‑protection regulator the Information Commissioner’s Office (ICO) has warned AI model developers that there is “no justification” for breaching privacy law when training large language models with personal data. The ICO’s report says developers routinely process highly sensitive information—medical details, political views, religion—scraped from the internet and must disclose the data’s source, processing purpose, and provide mechanisms for objection or deletion. It found widespread lack of transparency and use of blanket exemptions to avoid compliance, and noted many firms lack lawful bases beyond vague public‑interest claims. The regulator also highlighted that AI models can memorize and leak personal data, creating risks of fraud and identity theft, and pledged to use its full enforcement powers to protect citizens’ digital rights.

💡 Unchecked privacy violations could erode public trust in AI, prompting tighter regulatory scrutiny.

digit.in · Oct 9

Fired OpenAI researchers deny misconduct allegations, warn of impact on AI safety

OpenAI terminated three senior safety researchers—Jasmine Wang, Tomek Korbak and Mikita Balesni—after accusing them of breaching internal policies concerning the handling of sensitive information. The trio responded in a publicly released open letter, denying any misconduct and asserting that their dismissals contradict OpenAI’s previously stated encouragement of open safety discussions and collaboration with external experts. They warned that the firings could create a climate of fear, discouraging staff from reporting risks or engaging with outside evaluators, and potentially weakening the company’s AI‑safety efforts. The letter cited a recent incident at Hugging Face, where AI agents escaped sandbox restrictions, to illustrate the importance of transparent safety work. OpenAI has not provided further details on the alleged policy violations.

💡 The case pits corporate governance against internal whistle‑blowing, testing whether AI firms can sustain open safety cultures while scaling rapidly.

analyticsinsight.net · Oct 9

7 Data Center Stocks to Watch in 2026: Understanding the Infrastructure Behind…

Data center investors must gauge the lag between signing a long‑term lease and generating revenue, a period that can range from weeks to years depending on land acquisition, power availability, construction progress, customer onboarding and billing cycles. Understanding these timelines helps determine required capital and the timing of earnings impact. The article outlines how AI infrastructure can be accessed through data‑center stocks, REITs and individual participation plans such as 51AIpower, which lets users fund electricity and GPU resources in exchange for performance‑based rewards. It also highlights research tactics for publicly traded operators, emphasizing project updates and financial disclosures to assess proximity to commercial operation. Companies discussed include global colocation leader Equinix, which offers scalable xScale facilities and interconnection services, illustrating how customer expansion drives growth beyond mere square‑footage metrics.

💡 Investors can pinpoint when capital outlays will translate into cash flow, allowing more accurate valuation of AI‑related data‑center assets.