- Nex AI News
- Posts
- Former NSA head joins OpenAI board and safety committee
Former NSA head joins OpenAI board and safety committee
Nex AI News
Former NSA head joins OpenAI board and safety committee
Sunday - June 16 - 2024
Retired Gen. Paul Nakasone, former head of the NSA, will join OpenAI’s board of directors and its “security and safety” subcommittee.
This high-profile addition aims to address critics concerned about OpenAI's rapid advancements without adequate risk assessment.
Nakasone's extensive experience from the Army, U.S. Cyber Command, and NSA brings valuable expertise.
OpenAI, now a key AI provider for tech, government, defense, and major enterprises, benefits from his institutional knowledge and connections.
Brave Software is integrating its Leo chatbot with search results from the Brave Search API, available in its browser.
This aims to provide users with up-to-date information, such as the latest scores, contextual details while reading, and recent topics for social media posts.
📈 NVDA NVIDIA Corporation | $131.88 | +2.277 |
🔻 GME GameStop Corp. | $28.70 | -0.42 |
🔻 TSLA Tesla, Inc. | $178.01 | -4.46 |
🔻 SOFI SoFi Technologies, Inc. | $6.46 | -0.340 |
📈 SIRI Sirius XM Holdings Inc. | $2.6100 | +0.0700 |
Today’s newsletter :
Former NSA head joins OpenAI board and safety committee
Brave integrates its own search results with its Leo AI assistant
Meta pauses plans to train AI using European users’ data, bowing to regulatory pressure
How much does ChatGPT cost? Everything you need to know about OpenAI’s pricing plans
Apple joins the race to find an AI icon that makes sense
Technique improves the reasoning capabilities of large language models
As Google Targets AI Search Ads, It Could Learn a Lot From Bing
Reduce AI Hallucinations With This Neat Software Trick
Apple Intelligence Won’t Work on Hundreds of Millions of iPhones—but Maybe It Could
Microsoft postpones ‘Recall’, its controversial AI feature that recorded everything you do
Retired Gen. Paul Nakasone, former head of the NSA, will join OpenAI’s board of directors and its “security and safety” subcommittee.
This high-profile addition aims to address critics concerned about OpenAI's rapid advancements without adequate risk assessment.
Nakasone's extensive experience from the Army, U.S. Cyber Command, and NSA brings valuable expertise.
OpenAI, now a key AI provider for tech, government, defense, and major enterprises, benefits from his institutional knowledge and connections.
Nakasone aligns with OpenAI's mission, emphasizing data acquisition for surveillance and AI development.
His insights will help OpenAI enhance cybersecurity by detecting and responding to threats, targeting institutions like hospitals, schools, and financial institutions.
He joins the board’s safety and security committee, currently evaluating the company's processes and safeguards.
Brave Software is integrating its Leo chatbot with search results from the Brave Search API, available in its browser.
This aims to provide users with up-to-date information, such as the latest scores, contextual details while reading, and recent topics for social media posts.
Brave emphasizes privacy, not requiring logins or storing conversations on its servers, and anonymizing requests to protect user identity.
Leo Premium, at $14.99/month, offers higher rate limits and access to the latest models with unlinkable tokens for privacy.
Brave continues to enhance its AI capabilities, introducing AI-powered summarization and the Leo AI Assistant. Competing with browsers like Edge and Opera, Brave leverages its integrated browser and search stack to refine its offerings.
Meta has paused plans to train its AI systems using data from EU and U.K. users after pushback from the Irish Data Protection Commission (DPC) and the U.K.’s Information Commissioner’s Office (ICO).
The DPC welcomed Meta's decision, following intense engagement.
Despite using user-generated content in other markets, Meta faces GDPR obstacles in Europe.
Meta had planned to update its privacy policy to use public content from Facebook and Instagram for AI training, effective June 26, but this has been paused due to regulatory concerns.
OpenAI’s ChatGPT platform keeps expanding with new features. It now offers a GPT Store for AI-powered applications and services.
The new "memories" feature lets you save preferences for tailored chats, and an upgraded voice mode for real-time interaction is coming soon.
ChatGPT offers various plans:
Free: Access to GPT-4o, web-augmented responses, GPT Store, file uploads, and basic data analysis.
Plus ($20/month): Higher capacity, early access to new tools, and advanced data analysis.
Team ($25-$30/user/month): Collaborative features, custom apps, and admin tools.
Enterprise: Enhanced privacy, customization, and admin tools, starting at around $60/user/month.
Edu: Similar to Enterprise but tailored for educational institutions.
Nonprofits: Discounted rates for eligible organizations.
We’ll keep this guide updated with new plans and features.
News
What else is nex?
Apple joins the race to find an AI icon that makes sense
This week, Apple joined the AI icon race with Google, OpenAI, Anthropic, and Meta, unveiling a circular shape with seven loops to represent "Apple Intelligence."
No one knows what AI looks like, yet it must be represented in user interfaces. Early AI icons were robots, wizard hats, or magic wands, but these imply inhumanity or mystery.
Technique improves the reasoning capabilities of large language models
Large language models like ChatGPT excel at tasks like drafting legal briefs but struggle with numerical or symbolic reasoning.
MIT researchers propose Natural Language Embedded Programs (NLEPs) to address this, enabling models to generate Python code for queries.
As Google Targets AI Search Ads, It Could Learn a Lot From Bing
Paula Thompson, VP of client strategy at Optimal, has been helping Plunge sell cold plunge baths via ads on Google and Bing.
Recently, Bing's AI-driven ads shifted focus from direct purchases to educational content about cold plunging.
Generative AI tools, including ChatGPT, often produce false information, termed "AI hallucinations".
Developers are working to reduce these errors using a method called retrieval augmented generation (RAG).
RAG enhances prompts with data from a custom database, leading to more accurate responses.
Apple Intelligence Won’t Work on Hundreds of Millions of iPhones—but Maybe It Could
APPLE HAS ANNOUNCED its next era. Your iPhone, Mac, or iPad will soon be guided by AI, called Apple Intelligence, launching later this year.
It makes Siri smarter, rewrites emails, creates new emojis, and turns sketches into AI art.
Microsoft announced its "Copilot Plus PCs" tool, promising to give computers a "photographic memory" to understand users' actions and offer new AI tools.
Critics raised privacy and security concerns, arguing the recordings could be misused.
Reply