# Google Maps Can Call Businesses to Find Appointments—but Only in the U.S.

Canonical URL: https://superpowerdaily.com/videos/c4b1ee50-ca9a-4f9c-a485-afe074247d9b
Published: 2026-11-17T22:10:17.000Z
Episode date: 2026-11-17
Language: English
Duration: 11:26
Publisher: Superpower Daily
YouTube: https://www.youtube.com/watch?v=99i5aoDLokQ

Google Maps can now call local businesses to find appointment openings in the U.S., while its AI trip-planning features expand more broadly. OpenAI’s healthcare agents still require human oversight, the FBI warns that ARTEX is speeding up cyberattacks, and Audible’s AI previews spotlight reader praise—though whether they omit criticism remains unproven.

## Key takeaways

- Ask Maps adds editable, map-linked itineraries, but business
- Healthcare agents navigate disconnected systems, while human
- AI-assisted attacks compress response time; review app
- Audible mixes AI overviews with human recordings; whether

## Chapters

- 0:42 Maps Makes Plans: https://superpowerdaily.com/videos/c4b1ee50-ca9a-4f9c-a485-afe074247d9b?t=42
- 2:57 Healthcare Agents: https://superpowerdaily.com/videos/c4b1ee50-ca9a-4f9c-a485-afe074247d9b?t=177
- 5:53 Faster Cyberattacks: https://superpowerdaily.com/videos/c4b1ee50-ca9a-4f9c-a485-afe074247d9b?t=353
- 8:23 Audiobook Discovery: https://superpowerdaily.com/videos/c4b1ee50-ca9a-4f9c-a485-afe074247d9b?t=503

## Stories covered

- Google adds AI trip planning to Maps, with appointment finding limited to the U.S.: https://superpowerdaily.com/posts/google-adds-ai-trip-planning-to-maps-with-appointment-finding-limited-to-the-u-s
  Itineraries and saved-place searches are rolling out wherever Ask Maps is offered. Route tips and business-calling features start in the U.S.
- OpenAI brings computer-use agents to healthcare paperwork and clinical workflows: https://superpowerdaily.com/posts/openai-brings-computer-use-agents-to-healthcare-paperwork-and-clinical-workflows
  The agents can navigate medical records and insurer portals. OpenAI pairs access with company-run safety tests, customer deployment claims and configurable human review.
- FBI warns AI hacking tool ARTEX is accelerating cyberattacks worldwide: https://superpowerdaily.com/posts/fbi-warns-ai-hacking-tool-artex-is-accelerating-cyberattacks-worldwide
  The framework automates familiar intrusion techniques. Its attacks can trigger security alerts, but the FBI says their speed leaves defenders little time to respond.
- Audible adds AI audio previews that explain books and highlight reader praise: https://superpowerdaily.com/posts/audible-adds-ai-audio-previews-that-explain-books-and-highlight-reader-praise
  The discovery feature combines generated overviews with human recordings. Its emphasis on what readers liked raises a question about how much criticism listeners will hear.

## Transcript

Automatically transcribed from the narration.

[0:00] Hey everyone welcome to the Superpower Daily AI Briefing for November 17th 2026 I am so glad you're joining us today We've got a truly fascinating lineup of stories that illustrate exactly how artificial intelligence is shifting from merely answering our questions to actively taking action across the digital world It's a massive shift in how we interact with technology Here is our quick roadmap for today's briefing We're going to start with Google Maps rolling out new AI trip

[0:26] planning and automated appointment calling Then we'll look at OpenAI is deploying computer use agents to tackle health care paperwork Third an urgent FBI advisory on an AI hacking framework called Artex And finally how Audible is changing audiobook discovery Let's get right into it First up Google Maps AI trip planning New itinerary and appointment features So Google Maps is dramatically expanding what its AI assistant Ask Maps can actually do on Android and iOS This isn't just finding

[0:55] a spot on a map anymore It's about handing the whole logistical headache of planning right over to an AI agent powered by Gemini Picture this You want to plan a three day road trip filled with coastal hikes and oceanfront dining Step one you literally just ask Maps for the plan Step two the AI builds you a fully editable day by day schedule And because we're in Maps it drops numbered pins right onto your interface meaning your

[1:19] written schedule stays visually connected to your actual route It's so seamless And step three it's highly flexible It factors in live traffic and opening hours You can edit it on the go to add a spontaneous lunch stop Or if you like a bigger screen you can export the whole thing to Google Docs on your desktop Ask Maps will also search through your saved places checking live availability in real time which completely eliminates scrolling through endless bookmarks

[1:43] Now a crucial detail here is the geographic rollout Google is expanding Ask Maps to 30 European countries right now bringing its total reach to more than 200 countries and territories So the itinerary building and saved place searches those are available globally wherever Ask Maps is offered But and this is a big constraint the most agent like features are strictly US only for now In the United States you can actually have Ask Maps search online and actively

[2:09] call local businesses on your behalf to find open appointments We're talking beauty pet care auto repair home repair You tell it you need a nail appointment at 4 p m and it compares the available slots by checking online and calling around then presents you with the best options to make the final booking decision It's wild Also limited to the US at launch are the new know before you go route tips When you pull up directions the

[2:33] app proactively surfaces critical details in a really natural conversational way It warns you about parking availability at your destination points out EV charging stations along the way alerts you to expected weather conditions flags potential delays and even calculates tolls It pulls all of this from a massive database of over 300 million places and half a billion user reviews giving you the complete picture before you ever put the car in drive OpenAI agents in healthcare navigating medical

[2:59] records and insurer portals Let's see how this concept of AI taking action builds even further On November 16th OpenAI launched computer use agents for healthcare teams through its API and ChatGPT for healthcare platforms We all know healthcare administration is plagued by rigid disconnected systems right You've got an electronic health record over here an insurer's portal way over there and usually humans are the ones exhausted by copying and pasting data between them Well OpenAI's agents are designed

[3:28] to navigate across these interfaces clicking and typing just like a human In a clinical demo using synthetic data the workflow goes like this Step one the agent gathers evidence from the MRI order writes a clinician note and drafts a patient reply But crucially step three it prepares an authorization request and then pauses It waits for a human authorization specialist to review it before submission And we're already seeing this in production CBS Health reported one of their

[3:59] agents cleared a multi month document sending backlog over a single weekend Cigna is using them to collect member context from over 10 systems in mere seconds Another agent is handling system steps for tens of thousands of daily consultations But the vital point here isn't just that it works it's the cost OpenAI reported that on a Stanford benchmark called Health Admin Bench which covers complex stuff like prior authorizations and appeals their GPT 6 Astra model earned 94.3%

[4:27] of available rubric points at maximum reasoning effort But their Luna model it still earned an impressive 91.8% and it can perform these tasks for just 8.5 cents each That is a staggering 98% cheaper than using Astra making massive scale automation financially viable for hospital systems of literally any size Of course safety is the absolute priority with medical records OpenAI reports that in an evaluation of GPT 6 Astra handling 325 electronic health record inbox tasks tasks shaped

[4:58] entirely by real physician input reviewers rated 99% of them as safe No inaccuracies or omissions that could plausibly harm a patient That is a heavy critical metric Plus across 646 assessments physicians rated the AI's work as very or extremely helpful 93% of the time But despite these deeply impressive numbers the ultimate question for HIPAA compliant integration remains who gets the final click The key limitation here is that organizations absolutely must ensure human oversight remains in the

[5:30] loop Health systems can restrict permissions require explicit approvals before the agent submits any external paperwork and inspect step by step audit logs HIPAA compliance relies on eligible services covered by business associate agreements but it fundamentally depends on the safety guardrails organizations enforce in their own workflows The AI does the heavy lifting but a human must remain the final gatekeeper Alright shifting gears from automation that helps to automation that harms Story three FBI warns of Artex an

[6:00] AI hacking tool accelerating cyber attacks The FBI just issued an urgent flash advisory regarding an AI assisted penetration testing framework known as Artex Since at least August 2026 multiple threat groups have used this highly capable tool in worldwide intrusions spanning financial services government healthcare and defense Consider the timeline of a typical attack It brilliantly illustrates why defenders are so incredibly concerned about speed Artex doesn't have its own built in AI model Instead it connects to external

[6:30] large language models like those from Anthropic OpenAI or DeepSeek and coordinates specialized agents to execute commands At minute zero a target is selected Within just one hour these agents are automating reconnaissance rapidly finding vulnerabilities in overlooked systems like partner portals and exploiting them And in under three hours attackers are achieving elevated privileged access The FBI says Artex compresses the laborious work of chaining attack techniques together allowing hackers with minimal technical expertise to run complex devastating operations

[7:00] against multiple organizations at exactly the same time Because these attacks chain automated techniques so incredibly fast they are often noisy and trigger security alerts But defenders just don't have enough time to react before the damage is done So the FBI laid out critical defensive priorities Organizations need to aggressively secure internet facing apps They must rigorously review application permissions and importantly apply strict rate limits to sensitive APIs Defenders have to monitor for sudden bursts of automated requests

[7:29] look out for unexpected credential changes and investigate any large data transfers immediately We've already seen this in the real world In a recent attack on ConnectWise cloud backup Microsoft attributed the campaign to a China state aligned actor called Storm 040 They used Arnitech to accelerate password spraying and abuse permissions that customers had already granted to cloud applications And the truly frustrating reality Putting this genie back in the bottle is practically impossible The original developer known online

[7:58] as Autumn27 actually removed the Artex repository from GitHub back in October after reports of widespread abuse and made it closed source However the FBI explicitly warns that copies and community forks of the earlier open source version are already out in the wild They remain widely available to anyone looking The threat is permanently here and organizations just have to drastically adapt their defenses to this new velocity of AI driven attacks Let's wrap up today with a much

[8:24] warmer consumer story Story 4 Audible's AI audio previews that explain books and highlight reader praise Audible launched a new audio discovery feature on November 16th that completely changes how we choose audiobooks Right now it's limited to select English language titles in the US but it fundamentally redesigns the traditional preview Instead of just a random contextless five minute excerpt Audible is now serving up a highly structured three part pitch First there's an AI generated overview summarizing the

[8:56] book's core themes and highlighting what readers specifically liked drawing on public reviews and synopses Second it seamlessly transitions into archived human commentary usually a quick insightful snippet from a past interview with the author or an Audible editor And finally you get to hear the actual narrated sample from the book It's a really clever mix of synthetic summary and original human recording giving you a complete picture in just a couple of minutes totally saving you from scrolling

[9:23] through written reviews on your phone But the framing of these AI summaries is naturally sparking some debate TechCrunch raised a really great point If the AI is specifically prompted to summarize what readers liked does that mean optimizing for positive reader feedback will artificially limit your exposure to negative reviews or valid constructive criticism We don't know for sure just yet More testing of the system's actual outputs is definitely needed But it highlights a growing tension across our

[9:52] digital platforms When AI curates our discovery experience for maximum convenience who exactly decides what information gets left on the cutting room floor All right let's bring it all together Here are your briefing takeaways for today First Google Maps is expanding AskMaps itineraries globally to over 200 countries but restricting the most powerful agent capabilities like actively calling local businesses for appointments to the US Second OpenAI is proving AI agents can navigate messy healthcare portals at massive scale

[10:22] for just 8.5 cents but strict human oversight remains absolutely non negotiable Third the FBI's urgent warning about RTX shows AI is compressing entire cyber attack chains into under three hours leaving defenders dangerously little time to react And finally Audible's new three part audio previews offer a wildly convenient way to though prioritizing positive praise might mean missing out on honest critiques Looking closely at all four of these stories they all circle back to one massive undeniable shift

[10:54] Whether we're trusting an AI agent to book a simple nail appointment review a sensitive medical record summarize a 400 page book or defend a critical corporate network the sheer staggering velocity of these actions is forcing us to completely rethink our guardrails How does AI automation change the way we evaluate trust speed and security in our daily systems It's something we're all going to have to figure out very soon Read more at superpoweredaily com Stay curious

## Editorial standards

https://superpowerdaily.com/editorial-standards

This episode reports information available at publication time. Consult the linked stories and their reporting sources for context and subsequent updates.
