Microsoft COPILOT Update - It can go to Meetings for You and much more!

Microsoft previously introduced their AI assistant 365 copilot, and now there are new updates about it. Among the fresh features is one that allows the AI assistant to represent you in meetings. The inception of this assistant drew inspiration from Chat GPT, and it has been seamlessly integrated into Microsoft's suite of Office applications. Following a series of testing phases to ensure its proficiency, the assistant will become available to everyone starting November 1st. Going by the name Microsoft 365 Copilot, this assistant is designed to provide summaries of meetings held on teams for individuals who choose not to attend. Beyond that, it's crafted to assist in drafting emails, composing Word documents, designing spreadsheet graphs, and swiftly putting together PowerPoint presentations. Microsoft is optimistic that this tool will alleviate some mundane tasks. Although there's a growing concern among some individuals that it might phase out human workers, there's also an apprehension it might cause businesses to become overly reliant on AI powered assistants. Moreover, emerging regulations around AI mandate disclosure whenever content is AI generated and not human created. This AI assistant could potentially challenge these regulations. Both European and Chinese directives emphasize that individuals should be aware when they are interacting with AI as opposed to humans. Colette Stahlbaumer, the Spearhead of Microsoft 365, noted it's incumbent upon users to disclose they are utilizing copilot. She highlighted that even though the AI assistant can create responses, it's still up to people to use the tool in the right way. On the other hand, the European Union believes that the companies making these AI tools should make sure they are used responsibly. Before its wide scale launch, the writer had an opportunity to test copilot. Its foundation is built on the same technology that fuels ChatGPT, a creation of OpenAI. The demonstration was carried out on a Microsoft employee's laptop showcasing CoPilot's integration with an individual's account, accessing either personal or company data, all handled securely as per Microsoft, and not utilized for further training of the technology. The introductory experience revealed that Copilot could emerge as a useful tool and a formidable contender for individuals in office settings, particularly within companies striving to economize. The assistant adeptly condensed a lengthy email thread about a fictional product launch, proposed a response, and then tweaked the response to be more lengthy and informal as per the user's request, all without leaving any traces of AI intervention in the email. Copilot also showcased its prowess in crafting a multiple slide PowerPoint presentation in merely 43 seconds, deriving content from a word document. If the document contained images, it utilized them or sourced its own royalty free images. It even proffered a script to accompany the presentation. However, when tasked with making the presentation more colorful, it fell short and reverted to manual PowerPoint tools. In a practical Microsoft Teams meeting, Copilot demonstrated its ability to pinpoint discussion themes, offer summaries of different discussion threads, and even encapsulate a specific individual's contributions. It also delineated the pros and cons discussed in a disagreement represented in a chart format within a matter of seconds. Yet it's engineered not to critique individual performances in meetings, like identifying the most or least effective speaker, addressing the likelihood of people sidestepping meetings as Copilot can fill in for them. A Microsoft staffer quipped that meetings might morph into webinars. As of now, the technology can't discern between individuals sharing a device on teams unless verbally indicated. With a price tag of $30 per month, or roughly 25 pounds in the UK, CoPilot operates online and doesn't function offline. Critics argue that such technology could notably unsettle administrative roles. Carissa Velis, affiliated with Oxford University's Institute for Ethics in AI, voiced concerns over people becoming overly dependent on these tools. She highlighted potential pitfalls like tech breakdowns, hacking threats, glitches, or contentious new policies, especially if users become so dependent on the system that they feel disoriented without it. Alright, I hope you enjoyed the video and learned something new. Please like and subscribe for more AI content, and don't forget to hit the notification bell to stay updated. Thanks for watching and see you in the next one.