Tag: AWS

  • Slackbot Upgrade: New AI Features Tested in Major Slack Update

    Slackbot Upgrade: New AI Features Tested in Major Slack Update

    Key Takeaways

    1. Slack is enhancing its traditional bot into a full AI assistant, expanding its functionality beyond reminders and alerts.
    2. The new Slack AI assistant provides personalized plans, insights, and simplified searches within the workspace.
    3. It utilizes natural language processing for contextual responses and integrates with Microsoft Outlook and Google Calendar for scheduling.
    4. Traditional Slackbot features like custom commands and reminders will still be available alongside the new AI capabilities.
    5. The AI assistant operates within AWS’ virtual private cloud, ensuring data security and addressing concerns about data use in AI training.

    Slack is testing a significant upgrade that evolves its traditional bot into a full-fledged AI assistant, broadening its functions beyond just reminders and alerts. The enhanced Slackbot now has the ability to create personalized plans, gather insights from discussions and documents, and simplify intricate searches throughout the entire workspace.

    Major Changes Ahead

    Rob Seaman, the chief product officer at Slack, which is part of Salesforce, noted that this overhaul is essentially a total reconstruction. In an interview during the Salesforce Dreamforce 2025 event, he remarked, “The Slackbot we have today is quite basic.” He added, “We’ve really redesigned it as a unique AI partner.”

    As part of this pilot program, the Slack AI assistant is represented by an icon next to the search bar in the workspace. Clicking on it reveals a panel where users can ask Slackbot questions, like “What are my top tasks today?” or “Get the latest info on a project.”

    Contextual Responses

    The assistant utilizes conversations, documents, and threads within the workspace to offer contextualized answers that are personalized for each user.

    A standout feature of this upgrade is the natural language search. Rather than needing exact keywords, users can phrase their requests more freely. Slackbot is also capable of integrating with Microsoft Outlook and Google Calendar for meeting scheduling and checking availability.

    The update builds on Slack’s earlier AI features, which include automatic summarization and jargon interpretation. Seaman mentions that more nuanced AI enhancements will keep coming into play to “reduce user clicks.”

    Core Functions Remain

    It’s important to note that traditional Slackbot features—such as custom commands, automated messages, and basic reminders—will continue to be available.

    Slack stresses that the AI assistant operates within AWS’ virtual private cloud. Seaman assures, “no data leaves the firewall,” and none is utilized for training AI models, which addresses growing concerns regarding generative AI in business settings.

    While organizations can choose to forgo the AI Slackbot completely, individual users within a workspace do not have the option to disable it independently.

    The upgraded Slackbot has already been rolled out to 70,000 Salesforce employees internally. Slack is currently expanding its pilot to select customers, with plans for a wider release by year’s end.

    Source:
    Link

     

    Tags: , ,
  • AWS Outage Disrupted Smart Beds and Exposed Cloud Vulnerabilities

    AWS Outage Disrupted Smart Beds and Exposed Cloud Vulnerabilities

    Key Takeaways

    1. AWS downtime on October 20 disrupted various services, highlighting our reliance on cloud infrastructure.
    2. The issue originated from a malfunction in AWS’s DynamoDB in the US-East-1 region, causing widespread connectivity problems.
    3. Smart technology, like Eight Sleep’s Pod mattress, became inoperable due to dependence on cloud services during the outage.
    4. Eight Sleep quickly implemented a “Bluetooth low energy outage mode” to allow local control of their devices during internet failures.
    5. The incident underscores the need for local fallback systems in cloud-dependent technology to ensure functionality during outages.

    When Amazon Web Services (AWS) experienced downtime on October 20, the spotlight was on the impact it had on social media and streaming services. However, the effects were much broader. This incident highlighted how deeply intertwined our daily lives are with cloud infrastructure. In fact, some smart beds became nothing more than “wet bags” during the outage.

    Cause of the Disruption

    Amazon later identified the source of the issue. It originated in their US-East-1 region located in Northern Virginia, where their distributed database service, DynamoDB, malfunctioned, causing a chain reaction. This failure obstructed new online connections, which in turn disrupted EC2 virtual machines, Network Load Balancers, and associated backend APIs. Essentially, a vital data channel on the Internet became blocked, and the repercussions were felt worldwide.

    Impact on Smart Technology

    Among those affected was Eight Sleep’s expensive $7,000 Pod. This smart mattress utilizes water circulation and AI temperature mapping to enhance sleep quality. The downtime of AWS servers also impacted this sleep technology company.

    The Pod’s dependence on cloud-based machine learning and real-time biometric synchronization made it susceptible. Since its algorithms and user information are hosted on AWS servers, the outage caused it to be temporarily inoperable. One user even noted that one side of the bed reached a scorching 110°F.

    Quick Response from Eight Sleep

    Eight Sleep’s CEO, Matteo Franceschetti, addressed the issue on X and assured users that a fix was on the way. Within a day, the company introduced a Bluetooth low energy fallback they referred to as “outage mode.” This feature is a local control system that circumvents cloud APIs, allowing direct communication between the device and the app. Thus, users could adjust their beds via Bluetooth during Internet outages.

    Meanwhile, AWS engineers worked tirelessly to restore service. Once resolved, they rebooted essential systems to re-establish connectivity. Amazon issued a formal apology, labeling the outage “a significant event.” They committed to putting in place architectural safeguards to avert a future occurrence. However, this isn’t the first time such an incident has taken place.

    Historical Issues with AWS

    The US-East-1 region has consistently been AWS’ weak spot. Ironically, it is the oldest and busiest hub for the cloud company. Still, it was the center of outages in 2021, 2023, and 2024, impacting major services like Disney+, Slack, and Zoom.

    Each incident showcased how cascading failures within AWS’s internal dependencies exacerbated what began as minor system problems. Engineers have since cautioned that the extreme centralization of workloads in specific regions creates systemic risks. This could be addressed by more intelligent load distribution and regional redundancy.

    A Call for Change

    While the circumstances may seem humorous, they draw attention to a significant concern: the vulnerability of cloud-dependent technology. AWS, Google Cloud, and Microsoft Azure together manage the backend systems that a majority of connected devices rely on. These services support various gadgets, from thermostats to medical monitors. When they fail, so do the devices people think are self-sufficient.

    Consumer advocates argue that a ‘right-to-function’ standard should be established, necessitating that essential features continue to operate even during cloud service failures. Local fallback systems, such as on-device AI or Bluetooth controls, can maintain usability and reduce dependence on far-off data centers.

    The incident involving Eight Sleep serves as a wake-up call, emphasizing that the reliability of the “smart” revolution hinges on the servers that support it. Until hardware manufacturers prioritize local control, even the simple act of going to bed may rely on a data center remaining functional.

    Source:
    Link

     

    Tags: , ,