FROM OUR PARTNER
Get field-tested AI playbooks from leaders at Reddit, T-Mobile & Pure Storage
If 95% of enterprise AI pilots are failing, what separates the few that pass? Sept 10 to 18,
Glean's three-part series
brings in senior leaders to share how they moved from experiments to impact, covering the
full AI implementation lifecycle:
→ Foundation: T-Mobile's Principal Architect on building the foundation for scalable AI programs.
→ Adoption: Reddit's Lead Sales Enablement Manager on increasing AI adoption and rolling out AI agents.
→ Evaluation: Pure Storage's Digital Workplace Director on defining success metrics and ROI narrative.
🎁 Attend all three sessions to earn exclusive Glean Academy swag! 🎁
Enroll now
|
🔍 Apple plans an AI search engine for Siri
LINK
|
- Apple is developing an AI search feature for Siri, internally named "World Knowledge Answers", that will summarize web results using text, photos, video, and other multimedia elements.
- The company plans to power the new tool with a Google-developed model that will be hosted on Apple’s own secure Private Cloud Compute servers instead of on Google's cloud.
- Sources claim Apple also considered a partnership with Anthropic for its Claude models, but the firm reportedly asked for $1.5 billion a year, a higher price than what Google wanted.
|
📱 Instagram launches an iPad app after 15 years
LINK
|
- After nearly 15 years, Instagram has launched an official iPad app that is optimized for larger screens, replacing the stretched-out iPhone layout that people previously had to use.
- The new interface is built for the tablet, defaulting to the Reels screen and showing comments beside videos or letting you view a conversation next to your main chat list.
- Its Feed offers three viewing modes called “All,” “Friends,” and “Latest,” which let you see recommended posts, content from mutuals, or just a simple chronological timeline.
|
🤖 Tesla reveals new Optimus prototype with Grok AI
LINK
|
- A video on X reveals Tesla's next-generation Optimus prototype answering questions from Salesforce CEO Marc Benioff, demonstrating its early integration with the company's Grok artificial intelligence assistant.
- The new prototype has a fresh gold color and features hands that are much more detailed than previous versions, although they appear non-functional and similar to mannequin hands in the footage.
- Tesla previously said its next-generation hands would have actuators in the forearm operating the fingers through cables, a crucial improvement for performing both delicate and more imposing tasks.
|
⚖️ Scale AI sues former employee and rival Mercor
LINK
|
- Scale AI is suing competitor Mercor and former employee Eugene Ling, alleging he stole more than 100 confidential documents with customer strategies and proprietary information for the rival company.
- The suit claims Ling committed a breach of contract by trying to pitch Mercor's services to one of Scale's largest clients, identified only as "Customer A," before leaving his job.
- Mercor’s co-founder denies using any trade secrets but admits Ling possessed old files in a personal Google Drive, stating his company offered to destroy the documents before the lawsuit.
|
🤝 Atlassian acquires The Browser Company for $610 million
LINK
|
- Atlassian is buying The Browser Company for $610 million in cash, acquiring the new AI-focused Dia browser to connect its popular work apps like Jira and Confluence together.
- The deal gives Dia the distribution and sales team needed to compete, shifting its focus from personal tasks to exclusively serving individual users in a professional setting at their jobs.
- While Dia becomes a work-focused tool, the original Arc browser will only be maintained and not actively developed, with its future now uncertain under the new company ownership.
|
🌍 New AI turns photos into explorable 3D worlds
LINK
|
- Tencent's new AI, HunyuanWorld-Voyager, creates explorable video from a single image by generating 3D-consistent scenes that users can navigate using a defined camera path for movement.
- The model simultaneously produces RGB video and depth information, which can be converted into 3D point clouds for reconstruction, though it stops short of creating true 3D models.
- To learn this effect, Voyager was trained on over 100,000 video clips, including computer-generated scenes from Unreal Engine, teaching it to imitate camera motion in game environments.
|
|
