Welcome back! The biggest rivalry in mobile tech just got complicated. One giant has officially hired its competitor to fix its broken voice assistant. Another tech titan is promising to pay the electricity premiums for its data centers to stop communities from blocking construction. Meanwhile, your workplace chat app just turned into an autonomous worker that can run your other software. The cost of AI is no longer abstract. It shows up in energy bills, governance fights, and who gets to sit at the center of daily habits.
In today’s Generative AI Newsletter:
Apple and Google strike a Gemini deal to overhaul Siri
Microsoft pledges to shield communities from AI power costs
Salesforce turns Slackbot into a workplace agent
Windsurf rethinks coding with an AI-first editor
Latest Developments
The partnership between Apple and Google proves the severity of the Siri problem. Instead of relying solely on its own models, it has signed a multi-year deal with Google to bring Gemini into Siri, with a rollout planned later in 2026. While the terms of the deal are confidential, the implications are evident. The stakes are higher than making the voice assistant smarter because Apple has more than two billion active devices, which means whoever powers Siri can develop a daily habit.
Here is what the reporting points to:
Deal: Google called Gemini the most capable foundation for Apple’s foundation models.
Timing: The revamped Siri is expected to launch later this year, with a full rollout planned for 2026.
Overlap: Apple already routes some requests to ChatGPT, so this puts OpenAI in a backup position.
Backlash: Elon Musk called it an unreasonable concentration of power for Google.
This deal can enhance Siri's performance in daily interactions, especially with follow-up commands and longer queries. It also reinforces Apple’s dependency on Google, the same partner that pays Apple tens of billions of dollars to stay the default search engine. Users prioritize understanding what data is shared from their phone and what remains private over knowing model names. The broader AI race now looks less like a distribution deal, with Samsung already taking the Google route and Apple joining the club.
Special highlight from our network
Join Outskill’s live weekend workshop and enter 2026 with the most valuable skill in your toolkit.
What you’ll get:
16 hours of live expert-led training
Build AI agents and automate 50% of your work
Learn 10+ in-demand AI tools
$5000+ in free AI bonuses when you attend
Includes: the 2026 AI Survival Handbook, Prompt Bible, Monetization Roadmap, and a personal AI toolkit builder—all free when you show up.
📅 Live sessions: Saturday & Sunday, 10 AM–7 PM EST
🎓 Join 10M+ learners already using AI to earn more and work smarter
Microsoft is attempting to diffuse a nationwide wave of protests with its new Community-First AI Infrastructure plan. The initiative arrives as data center cancellations hit a record high across the U.S. and the White House warns against passing the tech industry's energy bill to American households. By promising to insulate residents from the soaring costs of AI's power appetite, Microsoft is reframing the massive expansion of server farms as a civic benefit rather than a resource drain.
Terms of the pledge:
Direct Negotiation: Microsoft will settle power costs directly with utilities to ensure that infrastructure upgrades never appear on residential electricity bills.
Subsidy Rejection: Vice Chair Brad Smith stated the company will refuse discounted power rates or taxpayer-funded subsidies for its new AI facilities.
Infrastructure Funding: The plan includes paying premium rates for grid expansions and funding local water conservation projects to offset high cooling demands.
Democratic Veto: Future buildouts now require formal community approval following the collapse of 25 major data center projects due to public backlash.
Across the U.S., residents are seeing electricity rates rise while new facilities draw enormous amounts of power and water. Data center activism now spans more than 20 states, and Trump has stated that Americans should not “pick up the tab” for AI power consumption. Microsoft’s plan does not reduce AI’s energy appetite. It attempts to redraw who pays for it. Whether regulators and residents believe those costs stay contained will determine how much AI infrastructure gets built at all.
Salesforce has rolled out a rebuilt version of Slackbot that operates as an AI agent inside Slack, capable of handling real workplace tasks instead of just answering questions. For Business+ and Enterprise+ customers, the update is now available. It fills in the data gaps between Microsoft Teams and Google Drive and echoes Salesforce's overarching objective of incorporating AI directly into enterprise tools rather than launching stand-alone products. CTO Parker Harris is proposing that the secret to AI domination will not be a new app but rather a comprehensive redesign of the interface that workers use eight hours a day.
What actually changed inside Slack:
Cross-Platform Agency: If granted permission, Slackbot can now pull data and trigger actions inside rival ecosystems like Google Workspace to consolidate workflows.
Internal Adoption: Salesforce claims the agent is its most successful internal tool launch to date, citing organic "unmandated" usage among its own global workforce.
Architecture Pivot: The system utilizes Generative AI to function as a curated agentic experience rather than a static list of trigger-based responses.
Roadmap Expansion: Future updates are slated to include voice capabilities and collaborative web browsing to move Slackbot closer to a full "Jarvis" for the office.
Salesforce CTO Parker Harris calls it a “super agent,” but the real change is structural. Slack is shifting from a communication layer to an execution layer. Salesforce is leveraging brand familiarity to mask a massive expansion of its data reach. This agent is designed to be the glue for a fragmented enterprise stack, but it also creates a central point of failure and a new layer of corporate surveillance. We are entering an age where the office chat app is where the work happens without you.
Windsurf is a coding editor built around an AI agent called Cascade. It can read your repo, understand file connections, and assist you make codebase changes. It lets you refactor, debug, test, and deploy tiny features without hand-editing ten files.
Core functions (and how to use them):
Repo refactors: Give Cascade a function name or module move. It updates imports and references across files to catch hidden usage.
Using context to fix bugs: Provide an error and the file you worked in. Ask it to find potential causes in linked files and suggest a testable repair.
Test generation for your project: Request function-specific unit tests. Tell it where your tests are and which framework you use to follow patterns.
File cleanup: Point it at a long script and ask it to split logic into smaller functions, remove duplication, and clarify names without changing behavior.
Repeatable processes: A workflow like “PR review” or “release prep” can be saved and executed later for consistent checks and suggestions.
Try this yourself:
Pick one annoying but small task in your current project: a rename, a duplicated helper, or missing tests. Open Windsurf and paste this into Cascade: “Refactor function_name for readability without changing output. Update all call sites across the repo. Then add 3 unit tests that cover edge cases, using the same style as existing tests in /tests.”
After testing, if errors occur, reply with: "Undo the last change; fix failing tests only. Keep the refactor."
Special highlight from our network
Clueso turns rough screen recordings, slide decks, and documents into polished videos. It writes the script, adds a natural-sounding voiceover, and generates transitions, effects, captions, and branding automatically. You stay in control at every step. With one click, you can translate your video into more than 50 languages.
More than 1,000 teams rely on Clueso every day. They use it for product demos, onboarding, marketing, and training.
From rough draft to ready-to-share video in one workflow.
