Microsoft is now the Copilot company, Chief Executive Satya Nadella declared this week at the company’s Ignite conference, not only spreading generative AI across its products but also announcing its first AI processor designs.
The flurry of news at Ignite was just one sign of how much gen AI is getting infused into every enterprise software company’s products. But it still may take awhile to become just as widespread in enterprises themselves, thanks to a raft of challenges in implementing it effectively and safely.
And not least, generative AI is providing a boost to supercomputing, as we found at the SC23 conference. It’s apparent there’s a lot of new thinking about the entire computing stack as AI increasingly defines new architectures for new kinds of workloads.
Analysts John Furrier and Dave Vellante discuss this and other news, zeroing in on the big AI chip battle, on their weekly theCUBE Podcast, out early today on YouTube. And this weekend, Vellante will run his latest not-to-be-missed deep-dive Breaking Analysis.
Here’s a sampling of the news this week from SiliconANGLE and beyond:
AI and the cloud
Some reality checks on generative AI:
You already knew this, but 2023 is the year of gen AI hype, says a report from Menlo Ventures out this week, called “The State of Generative AI in the Enterprise.” The point is, enterprises are still a long way from embracing it, the report says, though it also says 2024 will be the year of “real AI adoption.” And maybe it’s no surprise enterprises are taking some time to try to do it right, since, as Zeus Kerravala reports, Cisco says they’re not ready.
I’ve had doubts about the “fair use” argument by many AI companies trying to justify how they trained their models on all manner of content — like, whole books, websites, music catalogs and art collections — without asking first. So does Ed Newton-Rex, former VP of audio at Stability AI who makes a credible case against the argument in Why I just resigned from my job in generative AI. “Companies worth billions of dollars are, without permission, training generative AI models on creators’ works, which are then being used to create new content that in many cases can compete with the original works. I don’t see how this can be acceptable in a society that has set up the economics of the creative arts such that creators rely on copyright.” YouTube, it appears, might be taking a better approach that’s more cooperative with artists: YouTube lets Shorts creators use AI-powered tools to create music using famous artists’ styles
I finally got to reading Unbundling AI by Benedict Evans and it articulates some of my misgivings about chatbots, cool as they are — namely that they’re a bit of a return to the command line (too much work), and answers usually need to be checked (even more work). It remains to be seen how fixable it is given that it’s inherently a probabilistic model: As Evans puts it, this is not a database, so even basing them more on vetted and specific data might not entirely do the trick. But… it’s really early and tough to know how this technology will develop and for what. It still seems likely to have a huge impact — it’s no surprise OpenAI has had to slow signups for ChatGPT Plus — but I’m not sure we really know just how that will play out, except that what we’re predicting now probably will be incomplete at best.
It was all about AI at Microsoft Ignite
Microsoft debuts its first processors, a significant move that, somewhat surprisingly, doesn’t seem to bother its graphics processing unit partner Nvidia, whose CEO Jensen Huang once again joined Microsoft CEO Satya Nadella onstage. For his part, Nadella now calls Microsoft the Copilot company — glad that’s not a drinking game or we’d all be under the table — and it’s hard to envision a bigger bet for the software and cloud giant. A good bet, honestly. We had all the news here:
And more AI news
Gartner forecasts 20% cloud computing growth next year, driven of course by gen AI. Another take on that trend from the Wall Street Journal: Companies Tried to Spend Less on Cloud. Then AI Showed Up.
The AI model battle quickens: OpenAI Hustles to Beat Google to Launch ‘Multimodal’ LLM (from The Information)
A eulogy for coding in The New Yorker.
Even mo’ money — and by the way, artificial general intelligence at some point: OpenAI reveals new details about its AI development roadmap and fundraising plans
This didn’t get much attention, but it seems pretty interesting: Martian debuts novel AI model mapping technology for apps to leverage multiple LLMs at once Model routing is fundamentally about understanding how LLMs work and what makes them succeed or fail. “The better one understands these models, the more effectively you can route between them.”
Another example of the AI moving to where the data is: Dell partners with Hugging Face to help enterprises deploy generative AI on-premises
YouTube Co-Founder Hurley’s Stealth AI Startup Raises Money from A-Star Capital, Ron Conway (from The Information)
AI revs supercomputing
SiliconANGLE and theCUBE covered the big news at SC23, the supercomputing conference in Denver, from Nvidia, Intel, Dell and more, as supercomputing becomes an even bigger deal in the gen AI era. John Furrier dug deep into what’s coming next in computing architectures as gen AI changes the nature of applications. For an intriguing dive on that, check out his interview with Renu Raman, founder and CEO of Terizza, a stealth company rethinking the computing stack with AI models as the middleware layer. Here are more highlights from the event:
And more enterprise news
All that AI is going to require a lot of new data centers: Stack Infrastructure raises $290M in debt financing to build more data centers
Anything but the Nvidia GPUs you can’t get these days: Scaleway delivers cost-optimized Ampere Altra instances as an alternative for AI workloads
Hmm, this seems like a setback for a company that had claimed more momentum just a quarter ago. Big question: What does it say about enterprise spending in the next couple of quarters? Cisco offers light guidance as new product orders slow, sending its stock lower
A shot at Databricks and Snowflake? Onehouse open-sources its OneTable data tool with support from Google and Microsoft
And Databricks looks to one-up Microsoft’s Fabric announcement at Ignite: Databricks debuts AI-powered Data Intelligence Platform
A milestone for a bid to track and predict costs across clouds: FinOps Foundation debuts first major release of its FOCUS cloud billing standard
CoreView acquires Simeon Cloud to simplify Microsoft 365 setup and management, and oh, by the way, appoints a new CEO
Elsewhere around tech
You can’t walk this back, Elon: As Axios’ Dan Primack puts it: “No investment firms that backed Musk’s takeover of X, nor those that have backed his other ventures, made public statements of opposition or condemnation. Instead, they cowered.” At least there are some consequences.
So I finally won’t keep missing texts from my iPhone friends? Finally: Apple moves to make RCS messaging available on iPhones amid regulatory scrutiny
Oops: Google witness accidentally blurts out that Apple gets 36% cut of Safari deal Except then Google CEO Sundar Pichai confirmed it.
No car salesmen: Amazon is about to get into the business of selling cars
Well, that’s the height of chutzpah, as David Strom put it: Ransomware attacks now come with SEC breach complaints
Palo Alto Networks earnings top estimates, billings guidance misses stock falls 5% (from Investors Business Daily)
Comings and goings
Salesforce appoints a new CEO for Slack, Denise Dresser, formerly president of Accelerated Industries.
Alternative cloud provider Virtuozzo appoints Sergey Dobrovolsky, formerly vice president of cyber infrastructure at Acronis, its new chief technology officer.
Supercloud 5: The Battle for AI Supremacy Nov. 28-Dec. 1: SiliconANGLE and theCUBE will present insights into how the big cloud players as well as on-premises veterans are angling to take advantage of the huge AI opportunities, with coverage that includes onsite reporting and interviews from AWS re:Invent in Las Vegas, anchored in a live virtual event from our Palo Alto studio. Register here for the event.
HPE Discover Barcelona Nov. 29-30: TheCUBE will be there live and SiliconANGLE will have the news.
Image: Bing Image Creator; photo: Microsoft/livestream
Your vote of support is important to us and it helps us keep the content FREE.
One click below supports our mission to provide free, deep, and relevant content.
Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.