- Synthetic
- Posts
- Microsoft reveals Copilot+ PCs, Copilot Agents, Team Copilot, and real-time video translation
Microsoft reveals Copilot+ PCs, Copilot Agents, Team Copilot, and real-time video translation
Plus, Scarlett Johansson wrestles with OpenAI and Eric Schmidt warns on major AI risks
Steve Brown
May 23, 2024
In partnership with
This Week in AI
In a case of life imitating art, it has emerged that OpenAI’s Sam Altman approached actress Scarlett Johansson last year to ask if he could license her voice. Over a decade ago, Johansson provided the voice for the friendly AI, Samantha, featured in the 2013 hit movie, “Her," and Altman thought her voice would be “comforting to people.” Johansson declined his offer. Last week, she was shocked to hear a similar voice, “Sky,” used in public demos of GPT-4o. OpenAI has agreed to halt all use of the voice. The exponential improvement in AI capabilities has turned science fiction into science fact in a single decade. “Her” seems more prescient than ever.
Hear Scarlett Johansson’s statement on the matter here, as read by Sky. Oh, the irony 😀 And all great publicity for OpenAI.
Not to be outdone by Google’s I/O event last week, Microsoft held its annual Build event in Seattle, WA. Surprising nobody, Microsoft spent much of the three days discussing how it will integrate AI features into everything it does. Highlights include:
Copilot Agents - New autonomous agents built into Copilot will carry out a series of automated tasks without being prompted to do so. Custom copilots are also planned. Copilot agents will be available in Microsoft Copilot Studio later this year.
Team Copilot - Designed to be a new ‘team member’ that can facilitate meetings, manage agendas, track action items, summarize meeting outcomes, offer advice, and track and manage projects. Sounds similar to Google’s AI Teammate.
Phi-3-Vision - a new mini multimodal version of the Phi-3 model released last month, compact enough to run on a mobile device.
Real-time video translation on Microsoft Edge - a wow feature for Microsoft’s browser that will translate videos on YouTube, LinkedIn, Coursera, Reuters, CNBC, Bloomberg, and other sites into a user’s native language using subtitles and dubbing. Spanish to English and English to German, Hindi, Italian, Spanish, and Russian will be supported initially (coming soon), with additional language support to follow.
Copilot+ PCs - Microsoft branding for the emerging AI PC category that defines a minimum set of capabilities. To qualify, PCs must pack at least 40 TOPS (trillion operations per second) in a Neural Processing Unit and have all-day battery life. This high performance will power unique features, including Recall to easily find content you have viewed recently (a feature that has raised privacy concerns), Cocreator for near real-time image generation, and Live Captions to translate audio from over 40 languages into English. Qualcomm makes the first chip to qualify. In the second half of the year, Copilot+ PCs from Acer, Dell, Lenovo, and others will be launched based on qualifying chips from Intel and AMD.
Following a similar deal with Google, Reddit secured a content deal with OpenAI. The deal gives OpenAI access to real-time Reddit data via an API. As part of the deal, Reddit will develop new AI-powered tools for its users and moderators that use OpenAI technology.
Video: Copilot+ PCs
If you want to see what Microsoft promises for the new wave of Copilot+ PCs, look no further than their flashy new marketing video (1m 26s).
What Win32 was to graphical user interface, we believe the Windows Copilot runtime will be for AI
This week, we have a second video for you. Former Google CEO, Eric Schmidt, shares his perspective on the future of AI, autonomous agents, the role of open source, future risks, competition between China and The West, and much more. At times optimistic and often sobering, this is a must-view video for anyone thinking about the future of artificial intelligence and humanity. (20 mins)
AI Tech and Innovation
As GPT-4o and Gemini make old-school voice assistants Alexa, Google Assistant, and particularly Siri look well beyond their sell-by dates, Amazon is set to improve Alexa using generative AI technologies. Replacing the decade-old assistant later this year, the offering will NOT be part of Amazon’s Prime subscription service, though pricing has yet to be determined. Google Assistant and Siri are expected to get similar makeovers. Watch the WWDC 2024 keynote on June 10th to understand what Apple is working on in this area. iOS and macOS are expected to get significant AI-related upgrades with GPT-4o-like features.
Cybersecurity has long used machine learning and artificial intelligence. AI-powered tools scale efforts to monitor many attack surfaces and automate routine tasks. Generative AI democratizes cybersecurity access, making it easier for security professionals to understand their security posture and find issues more quickly. New AI-powered cybersecurity tools mitigate risks more effectively and deliver hyper-automation to augment human security analysts and empower them to handle the ever-expanding threat landscape.
AI Insights
Most internet access, in my view, will be by agents; not by human beings. That might have some implications on how content on the internet needs to change
One thing’s for sure: to build out the AI capabilities that businesses and consumers will want in the future, we will need a lot of high-end chips. These will be built on the latest manufacturing nodes inside factories that don’t yet exist. The CHIPS and Science Act was a bold move by the U.S. government to ensure that some of these chips will be built on U.S. soil. In 2022, the United States made precisely 0% of the world’s most advanced chips (found inside the latest smartphones and AI data centers). Thanks to the CHIPS and Science Act, we may see that percentage jump to 28 percent by 2032, though challenges still remain.
At least in their first instantiations, early AI-focused gadgets like the Rabbit R1 and Humane Pin have failed. Despite their tagline, “Things are looking up”, Humane is reported to be looking for a buyer only a month after they launched their first product. However, these products questioned the role AI could and should play as the future interface for digital devices in our busy lives. Smartphones are amazing. But for many of us, they are a complex blur of hundreds of apps, each vying for our attention when we are just trying to get something done. Will the big mobile OS vendors use AI to simplify and focus the user experience on what we need right now? Could the app model slowly die away as a result? Synthetic thinks so, and so does the writer of this piece.
A study by Washington State University, published in the International Journal of Contemporary Hospitality Management, surveyed over 620 lodging and food service employees and concluded that technophobia increases worker stress and a sense of job insecurity, making them more likely to leave their jobs and frustrating the efforts of companies trying to alleviate worker shortages with automation technology. People with experience working alongside robotic technology are even more likely to feel this way. Turnover in hospitality is high, and the sector has had trouble attracting workers in the post-COVID era, leading many employers to turn to robotic technology. While not explicitly mentioned in the report, robot phobia will likely apply to manufacturing, construction, mining, warehousing, and other physical labor sectors.
Toolkit for the Future
Learn how to become an “Intelligent Investor.”
Warren Buffett says great investors read 8 hours per day. What if you only have 5 minutes a day? Then, read Value Investor Daily.
Every week, it covers:
Value stock ideas - today’s biggest value opportunities 📈
Principles of investing - timeless lessons from top value investors 💰
Investing resources - investor tools and hidden gems 🔎
You’ll save time and energy and become a smarter investor in just minutes daily–free! 👇
This versatile AI assistant helps content creators write text, edit content, change writing tone, create images, generate audio, easily translate text into other languages, quickly summarize hours-long YouTube videos, generate memes, and more. Free to try.
Extract data from websites and place it in a spreadsheet that fills itself and notifies you of any changes. Pre-built robots make it a breeze to get started. Extract whatever you need: job listings, property details, TikTok videos, hotel info, or competitor pricing. Free to try. No coding needed.
A new set of generative AI-powered tools to transform images, remove backgrounds and watermarks, create backgrounds, and visualize products. Deliver one-of-a-kind visual experiences and better engagement on the web. Free to try.
Keep your sales team focused on customers during meetings. Laxis captures attendee comments verbatim and flags items for follow-up. It auto-generates meeting summaries and follow-up emails in seconds, quickly identifying customer requirements, pain points, and action items. Free to try.
AI extracts information from your receipts and organizes it in an easy-to-search database alongside IRS-approved, secure image scans. Capture receipts on the go📱, forward receipts 📧, use the web portal, or mail in a pile of physical receipts🧾. Try it free for 30 days.