At Google’s annual I/O developer conference last week, the company unveiled a slew of new artificial intelligence features and updates. Many of these announcements were met with tepid reactions, often because they felt like flashy demonstrations rather than tools that would genuinely improve daily life. However, one Gemini-powered capability has proven itself in the days following the event: the Daily Brief. This feature, while humble in concept, delivers a genuinely useful daily to-do list generated from a user’s Google Workspace data.
The Daily Brief lives in the left sidebar of the Gemini app and relies on Google’s Personal Intelligence system. Personal Intelligence connects core services like Gmail, Google Calendar, and Google Drive to Gemini, enabling the AI to understand a user’s personal context. To activate Daily Brief, users must first enable Personal Intelligence through their profile icon in the Gemini app, then navigate to the Personal Intelligence settings. Once enabled, the Daily Brief appears as a panel that updates each morning.
The author, who has been a vocal critic of Personal Intelligence in the past due to its often gimmicky and intrusive suggestions (e.g., “Since you’re already renovating your New York apartment…”), found Daily Brief to be a refreshing change. The feature avoids fluff and gets straight to the point. After a smarmy but friendly greeting like “Hey Ben here’s what today looks like” punctuated by an oversized emoji, the Daily Brief presents a series of bulleted lists. These lists contain action items extracted from Gmail, calendar events, and recent Gemini chats. Users can see where they need to go, whom they need to reply to, what appointments require preparation, and which chores need to be tackled.
For someone who appreciates structured to-do lists, the standout usability factor lies in the action buttons that accompany each item. Event invitations include an “add to calendar” button, while chores and tasks provide icons that open the relevant emails. The most satisfying feature, however, is the “Mark complete” button hidden within the three-dot “overflow” menu. There is also a Chat button that spawns a Gemini conversation about the item. Tapping “Mark complete” grays out the associated to-do, delivering a dopamine hit of accomplishment.
After using Daily Brief for approximately one week, the author reports that it has been mostly accurate in capturing key appointments and tasks. It has even highlighted important items that might otherwise have slipped through the cracks. The feature is available to subscribers of Google’s $20-per-month AI Pro plan, as well as the higher-tier AI Ultra plan. Users who are uncomfortable with Gemini scanning their Gmail can easily disable the feature. For those who are willing to try it, Daily Brief offers a concrete example of how AI can enhance productivity without being overbearing.
More in AI this week
Beyond Google’s Daily Brief, several other AI developments made headlines. In a historic move, Pope Leo XIV issued his first encyclical, addressing the promise and perils of artificial intelligence. He emphasized that AI can be a valuable tool but must be used with vigilance and thoughtfulness. For everyday users, this means keeping their own neurons active while working with AI rather than switching them off.
Claude, the AI assistant by Anthropic, has been receiving complaints from users who say it nags them to take breaks and sleep, interrupting night owls and early birds alike. Anthropic has acknowledged the issue and is looking into it.
Apple is reportedly on the verge of releasing a long-promised AI update for Siri. The update is expected to allow Siri to access personal data, analyze on-screen content, add calendar events, launch apps, and search the web. This move, reported by Bloomberg, could finally bring Siri up to speed with competitors like Google Assistant and Amazon Alexa.
There is also speculation about whether corporate America is having second thoughts about aggressively investing in AI after burning through budgets. While some analysts argue that the hype is cooling, others maintain that spending on AI training and inference will continue to grow. A blog post by Simon Willison explores the sentiment around “tokenmaxxing” and whether companies are reevaluating their ROI.
Finally, the concept of recursive self-improvement (RSI) is gaining attention in AI research circles. RSI refers to an AI system that can automatically and continuously improve itself without human intervention. This idea, while still theoretical, raises significant concerns about control and safety. TechCrunch recently covered the emerging discussion around RSI.
Prompt of the week: The “wrong question” meta-prompt
Many users find that AI responses are not as useful as they hoped, often because the initial question was poorly framed. A too-specific prompt can result in a narrow answer that misses the broader context, while a too-vague prompt yields a wishy-washy reply. Sometimes, the problem is that the user is asking the AI to solve a problem that does not actually need fixing.
A solution is the “wrong question” meta-prompt. This technique involves asking the AI to interrogate the original prompt: to challenge its preconceptions, structure, and framing. The goal is to bridge the gap between what the user is literally asking and what they truly need. For example, before presenting the final prompt, the user can instruct the AI to first analyze the request for hidden assumptions or missing context, then suggest a better formulation. This meta-cognitive approach helps refine the input and yields more accurate and insightful outputs.
That’s all for now. Thank you for reading. If you enjoyed this article, consider subscribing to the newsletter for more weekly insights on using AI in everyday life.
Source: PCWorld News