This article discusses how to run large language models like Gemma 4 locally on Android or iOS devices using the Google AI Edge Gallery. By leveraging MediaPipe and TensorFlow Lite, developers can build applications that perform tasks such as text classification and image captioning without an internet connection. This approach enhances user privacy by keeping all data on-device and enables functionality in areas with limited connectivity.
* Local execution of large language models via edge infrastructure
* Privacy benefits through on-device processing
* Offline capabilities for various machine learning tasks
* Open source access to demo code via GitHub
* Local execution of large language models via edge infrastructure
* Privacy benefits through on-device processing
* Offline capabilities for various machine learning tasks
* Open source access to demo code via GitHub
While cloud-based AI models are more powerful, running small language models locally on a smartphone offers unique advantages in privacy and practicality. This article explores how on-device LLM can be used for tasks that don't require massive computing power but benefit from being offline or private. Key use cases include:
* Using it as a private thinking partner for personal questions.
* Organizing messy, unstructured notes and brain dumps.
* Performing quick code logic checks and debugging snippets while away from a computer.
* Acting as a low-pressure language tutor that works without an internet connection.
* Using multimodal capabilities to analyze images like whiteboards or product labels via the phone camera.
* Using it as a private thinking partner for personal questions.
* Organizing messy, unstructured notes and brain dumps.
* Performing quick code logic checks and debugging snippets while away from a computer.
* Acting as a low-pressure language tutor that works without an internet connection.
* Using multimodal capabilities to analyze images like whiteboards or product labels via the phone camera.
**General Security Practices (Apply to All Users):**
* **Use End-to-End Encryption:** Adopt a secure messaging app like Signal (compatible with iOS & Android).
* **Secure Messaging App Hygiene:**
* Be wary of social engineering attempts (account compromise scams).
* Verify group invitations through separate channels.
* Be suspicious of unexpected security alerts *within* the app.
* Enable message expiration features (check organizational policies first).
* Regularly review and remove unintended linked devices.
* **Enable FIDO Authentication:** Use phishing-resistant MFA (hardware keys like Yubico/Google Titan preferred, passkeys acceptable) for valuable accounts (email, social media, etc.). Disable less secure MFA methods after enabling FIDO.
* **Migrate Away from SMS MFA:** Do *not* use SMS for two-factor authentication. Use authenticator apps (Google Authenticator, Microsoft Authenticator, Authy) as a better alternative, but FIDO is strongest. Disable SMS after enabling an authenticator app.
* **Use a Password Manager:** Store all passwords in a password manager (Apple Passwords, LastPass, 1Password, etc.). Use a strong passphrase to protect the vault. Regularly update passwords with the manager.
* **Set a Telco PIN:** Add a PIN/passcode to your mobile account with your provider to prevent SIM-swapping. Also, change your mobile account password.
* **Regularly Update Software:** Update operating systems and apps weekly; enable auto-updates.
* **Upgrade Hardware:** Use the latest hardware version from your phone manufacturer for optimal security features.
* **Avoid Personal VPNs:** They can increase the attack surface. Use organization-provided VPNs if required.
**iPhone-Specific:**
* **Enable Lockdown Mode:** Reduces the attack surface.
* **Disable "Send as Text Message":** Ensures iMessage (end-to-end encrypted) is used when available.
* **Protect DNS Queries:** Use Apple iCloud Private Relay or encrypted DNS services (Cloudflare, Google, Quad9).
* **Enroll in iCloud Private Relay:** Enhances privacy and security for Safari browsing.
* **Review App Permissions:** Restrict access to sensitive data (location, camera, microphone).
**Android-Specific:**
* **Prioritize Secure Manufacturers:** Choose Android phones from manufacturers with strong security track records and long-term update commitments (check Android Enterprise Recommended).
* **Use RCS with Encryption:** Only use RCS (Rich Communication Services) if end-to-end encryption is enabled (Google Messages).
* **Configure Android Private DNS:** Use a trusted DNS resolver (Cloudflare, Google, Quad9).
* **Enable Secure Connections in Chrome:** Ensure all website connections default to HTTPS.
* **Enable Enhanced Protection in Chrome:** Provides an extra layer of protection against malicious websites.
* **Enable Google Play Protect:** Regularly review app scans for threats. Be cautious with third-party app stores.
* **Review App Permissions:** Restrict access to sensitive data.
* **Use End-to-End Encryption:** Adopt a secure messaging app like Signal (compatible with iOS & Android).
* **Secure Messaging App Hygiene:**
* Be wary of social engineering attempts (account compromise scams).
* Verify group invitations through separate channels.
* Be suspicious of unexpected security alerts *within* the app.
* Enable message expiration features (check organizational policies first).
* Regularly review and remove unintended linked devices.
* **Enable FIDO Authentication:** Use phishing-resistant MFA (hardware keys like Yubico/Google Titan preferred, passkeys acceptable) for valuable accounts (email, social media, etc.). Disable less secure MFA methods after enabling FIDO.
* **Migrate Away from SMS MFA:** Do *not* use SMS for two-factor authentication. Use authenticator apps (Google Authenticator, Microsoft Authenticator, Authy) as a better alternative, but FIDO is strongest. Disable SMS after enabling an authenticator app.
* **Use a Password Manager:** Store all passwords in a password manager (Apple Passwords, LastPass, 1Password, etc.). Use a strong passphrase to protect the vault. Regularly update passwords with the manager.
* **Set a Telco PIN:** Add a PIN/passcode to your mobile account with your provider to prevent SIM-swapping. Also, change your mobile account password.
* **Regularly Update Software:** Update operating systems and apps weekly; enable auto-updates.
* **Upgrade Hardware:** Use the latest hardware version from your phone manufacturer for optimal security features.
* **Avoid Personal VPNs:** They can increase the attack surface. Use organization-provided VPNs if required.
**iPhone-Specific:**
* **Enable Lockdown Mode:** Reduces the attack surface.
* **Disable "Send as Text Message":** Ensures iMessage (end-to-end encrypted) is used when available.
* **Protect DNS Queries:** Use Apple iCloud Private Relay or encrypted DNS services (Cloudflare, Google, Quad9).
* **Enroll in iCloud Private Relay:** Enhances privacy and security for Safari browsing.
* **Review App Permissions:** Restrict access to sensitive data (location, camera, microphone).
**Android-Specific:**
* **Prioritize Secure Manufacturers:** Choose Android phones from manufacturers with strong security track records and long-term update commitments (check Android Enterprise Recommended).
* **Use RCS with Encryption:** Only use RCS (Rich Communication Services) if end-to-end encryption is enabled (Google Messages).
* **Configure Android Private DNS:** Use a trusted DNS resolver (Cloudflare, Google, Quad9).
* **Enable Secure Connections in Chrome:** Ensure all website connections default to HTTPS.
* **Enable Enhanced Protection in Chrome:** Provides an extra layer of protection against malicious websites.
* **Enable Google Play Protect:** Regularly review app scans for threats. Be cautious with third-party app stores.
* **Review App Permissions:** Restrict access to sensitive data.
The TC specifies a common protocol, framework and interfaces for interactions between AI agents using natural language while supporting multiple modalities.
The This framework will also facilitate communication between non-AI systems (e.g., clients on phones) and AI agents, as well as interactions between multiple AI agents.
The This framework will also facilitate communication between non-AI systems (e.g., clients on phones) and AI agents, as well as interactions between multiple AI agents.
Some 58% of all email opens it examined occurred on mobile phones or tablets (45% phones, 13% tablets) in 2Q16, a 14% increase from a year earlier.
Although desktop computers accounted for only 42% of email opens, they were responsible for 49% of email-generated revenue tracked by Experian.
Most 2Q16 email opens occurred on mobile phones or tablets for all industries except business products/services:
Although desktop computers accounted for only 42% of email opens, they were responsible for 49% of email-generated revenue tracked by Experian.
Most 2Q16 email opens occurred on mobile phones or tablets for all industries except business products/services:
