As summer effectively comes to an end, enterprise tech companies continued to forget August is supposed to be for vacations — but that’s not such a bad thing.

This past week Google mounted a renewed offensive in artificial intelligence at its Next cloud event, sharpening its message and its product lines alike at a packed three-day conference in San Francisco. Meantime, leading enterprise companies from Dell Technologies and Nutanix to Salesforce and MongoDB turned in surprisingly good earnings reports — Dell’s stock rose a stunning 21% on Friday — indicating that even though enterprise spending remains depressed, customers are upping spending where it counts: on generative AI, for one.

There was a lot of other news this week as well, including another potential near-term initial public offering reported today with data protection and cybersecurity firm Rubrik, adding to Arm, Instacart and others soon to float. So it looks like it could be an even busier fall.

You can get a deeper dive into the implications of all of Google’s announcements at Next, enterprise earnings and more on today’s theCUBE Pod podcast with SiliconANGLE analysts John Furrier and Dave Vellante. And don’t miss the latest installment Saturday of Vellante’s weekly Breaking Analysis deep dives, this one on how Google’s going all in on the AI cloud.

First, check out our latest Cloud Special Report, including extensive written and video coverage of Google Cloud Next.

Cloud Special Report

Cloud giants eye a potential windfall in AI at the network edge Another big market they’re going to swallow up.

Breaking Analysis: Snowflake has momentum with AWS and Microsoft – and why Google may not be Next

Google Cloud Next

First, a few thoughts after Next:

* Generative AI is breaking on-premises data centers: Drew Bradstock, Google Cloud’s senior director of product for Kubernetes and serverless, told me in an interview at Next that one big hedge fund was doing all its workloads on-premises. “AI broke their capacity model,” Bradstock said, so it had to move certain workloads very quickly to the cloud.

* Gen AI isn’t just for the latest whizzy apps but for large enterprises looking to migrate apps to the cloud, Gabe Monroy, vice president of developer experience for Google Cloud, told me. “It’s not about speeding up creation of the latest new app,” he said. “It’s how they modernize their legacy.”

* That said, SiliconANGLE’s John Furrier has the distinct sense that startups today may be looking first to Google for gen AI help on their apps rather than Amazon Web Services, which built itself on startups back in the Web 2.0 era. AWS as legacy? Say it ain’t so…. But as Google Cloud CEO Thomas Kurian (pictured) pointed out for not the first time at Next, more than 70% of generative AI unicorns are Google Cloud customers, including AI21 Labs, Runway AI, Typeface, Anthropic and Character.AI. And Next, more than anything an event for developers, was packed.

* At the same time, large enterprises need a lot of help. “Gen AI has the potential to go way over the heads of many of our enterprise customers,” Monroy said. That’s the reality behind the just-announced Jump Start Solutions, Google’s stab at providing readymade solutions for common use cases.

* Not all the AI capabilities Google announced are here yet. For example, Duet AI in Google Cloud is only in preview until later this year. But they’re coming fast, since Google has no intention of remaining behind Microsoft and OpenAI in AI capabilities, especially with at least a couple of decades of its own AI work, and seven years as an “AI-first” company, as Alphabet and Google CEO Sundar Pichai pointed out in his brief keynote Tuesday. “If you really think about what’s required around generative AI, that heritage is going to accelerate Google’s transformation,” said Matt Lacey, global chief commercial officer of Deloitte Consulting’s Alphabet Google alliance. Given that it still badly trails AWS and Microsoft in cloud, Google needs to take full advantage of that if it’s to have any hope of moving closer the cloud leaders.

* Despite the ongoing battle between cloud providers and on-premises providers — which the cloud guys have been winning for awhile, of course — hyperscalers need the likes of Dell Technologies and Hewlett Packard Enterprise. That’s because a lot of AI inference, the running of those models, will be done at the network edge, where those companies gear sits. “We’re counting on our partners to provide scale,” Sachin Gupta, vice president and general manager of Google Cloud’s infrastructure and solutions group, told me in an exclusive interview. “They also have global reach.” And not least, a lot of AI-friendly graphics processing units in that gear.

* Google told a simpler AI story at Next, after a few years in which its efforts looked a bit scattered, at least until recently. “Google has simplified its branding – with Duet AI as an all-in-app AI assistant and Vertex AI for AI builders, said Gopal Srinivasan, Deloitte’s Alphabet Google Alliance Generative AI Leader. “They’ve taken everything they’ve had before and totally simplified it into these two sets of solutions and what that will do is only make it easier for customers to wrap their arms around the capabilities they can apply to their enterprises.” And it’s not just branding. Google has the AI chops.

* The consensus so far is that gen AI will give the bad guys a leg up on new paths for cyberattacks before the defenders can leverage the technology. Jeff Reed, vice president of product for cloud security at Google Cloud, told me he’s not so sure about that. “I haven’t seen evidence of that so far,” he said, and Google, with its threat tracking unit Mandiant as well as Chronicle, is certainly in a position to know. In fact, Reed thinks large language models can find more security issues than the common method of sandboxing, not to mention helping solve the huge cybersecurity talent shortage by spreading knowledge to more people more quickly.

* Term of the week: vector embeddings. We’ll be hearing even on it, since it’s key to generative AI, in coming weeks and months.

Vellante’s Breaking Analysis goes a lot deeper into all the implications of Google’s AI cloud push. And here’s SiliconANGLE’s extensive coverage of the news and the event:

Out with the old, in with the new: AI shapes the narrative for Google Cloud Next

Google expands Vertex AI with new models and enterprise-ready tooling

Google applies AI to speed threat intelligence and tighten Workspace security

Google Cloud’s generative AI tech to power dozens of partners’ technologies

Google Cloud gives analytics teams tools to bring AI to data with BigQuery Studio and Duet AI

Google Cloud debuts its most powerful, AI-optimized VMs yet

Google DeepMind unveils tool to watermark and detect AI-generated images

From legacy land to cloud frontier: How Google Cloud is shaping the future of cloud computing — theCUBE analysis

AI moves by Google Cloud raise interest in what AWS will deliver next

Google Cloud Next final analysis: Transformative shifts and a cloud showdown

And check out all the rest of our video interviews and analysis on theCUBE, SiliconANGLE’s livestreaming video studio, onsite from San Francisco’s Moscone Center. (* Disclosure: Google sponsored some interviews, which are marked as such, but not our editorial coverage and analysis.)

Enterprise pulse

In late-breaking news Friday, another potential IPO: Rubrik could reportedly go public by year’s end

New funding round could raise Databricks‘ value to more than $43B

Sometimes antitrust more or less does what it’s supposed to do: Microsoft to unbundle Teams from Office in Europe amid antitrust probe

Intel looks to keep its lead in the data center as it trails badly in AI-oriented chips: Intel unveils next-gen Sierra Forest and Granite Rapids Xeon CPUs for data center workloads

A somewhat rare but typically small acquisition by Amazon, but since it’s for developers, it also may provide more bang for the buck: Amazon acquires Fig to enhance the command line for developers

What does every database need in the generative AI era? Embeddings. Case in point: Pivotmaster Rockset reels in $44M for its real-time analytics database

Earnings: Enterprise ups and downs — but mostly ups:

Software-defined (and AI) FTW: Dell, Nutanix continue the string of recent upbeat earnings news Dell’s stock jumps 21% on Friday, while Nutanix’s shares rise 12%.

Broadcom shares fall on lower-than-expected outlook, while its soon-to-be-acquired VMware more or less meets forecasts

MongoDB smashes earnings forecast and outlook as new AI-powered features take off

Salesforce shrugs off economic pressure, beating expectations and raising full-year guidance

Cybersecurity looks up, as CrowdStrike beats, Okta impresses again, and so does possibly acquisition-bound (or not) SentinelOne: CrowdStrike and Okta shares rise on solid earnings and revenue beats and SentinelOne shares rise as cybersecurity company tops forecast But IronNet, founded by former National Security Agency chief Keith Alexander, looks down for the count a year after it hit the skids.

Surprise earnings and revenue beat sends Pure Storage’s stock higher after-hours

Subscription customer growth prompts Elastic share surge in late trading — and on Friday, when they’re up 20%.

Shares of HPE and HP fall on soft guidance as enterprise spending lags

Box‘s stock falls as it reports narrow earnings beat and weak guidance

Roundup: HashiCorp and PagerDuty beat but not by enough and shares of both fall. Samsara beats and its shares rise 13% on Friday.

AI beat

Is the trough of disillusionment already coming soon in generative AI? Man, that Hype Cycle is in overdrive: AI Startup Buzz Is Facing a Reality Check But others are still getting big checks: OpenAI rival AI21 Labs raises $155M in funding

OpenAI goes enterprise: OpenAI launches ChatGPT Enterprise with bevy of cybersecurity features

And it’s already making a boatload of revenue (if not profit): Report: OpenAI on track to generate $1B+ in annual revenue

AI goes to Washington: Tech chiefs will meet in Washington next month to discuss AI regulations

Will gen AI kill no-code tools? It may have just ended AWS’ Honeycode cloud app-building service.

“Chicken Tikka Masala from Zareen’s, stat!” DoorDash debuts AI-powered voice ordering service

The issue of training data will continue to be contentious: In what could be a landmark case, Open AI lawyers motion to dismiss authors’ copyright claims Websites that gave up so much to Google in the past couple of decades don’t want to get fooled again.

Related food for thought from the guy who’s literally writing the book about Luddites: Actors and writers are Luddites in their stance about gen AI — but in the often misunderstood sense that they didn’t oppose technology, only the use of it against them by their bosses: Column: What Stephen King — and nearly everyone else — gets wrong about AI and the Luddites

At least they’re not copter pilots: AI quadcopter autonomous drone in Switzerland leaves champion human racers in the dust

Using AI to fix current and potential problems with AI: Activision partners with Modulate to combat in-game toxicity with AI-powered voice chat moderation and Meta releases FACET dataset for evaluating AI fairness

An interesting case study by Paul Gillin on how data structures need to be set up in advance to leverage gen AI: To accelerate business insights, a clinical lab turns to a data fabric

Cyber beat

David Strom looks into how gen AI could bolster an old cybersecurity method: Google’s new use of generative AI could boost ‘fuzzing,’ a longtime cybersecurity defense tool

Strom digs into another growing avenue of attacks: Phishing as a service continues to plague business users

Malwarebytes lays off 100+ workers ahead of planned company split

North Korea Lazarus Group beefs up its malware attacks once again

Microsoft once again revokes VeriSign digital certificates: Here’s why it matters

Multinational task force takes down prolific Qakbot malware and botnet operation

Meta removes thousands of accounts linked to Chinese government propaganda campaign

What’s coming next

Earnings next week:

Sept. 5: Zscaler, GitLab and Asana

Sept. 6: UiPath, Couchbase and C3.ai

Sept. 7: Secureworks, DocuSign and Smartsheet

Photo: Robert Hof/SiliconANGLE