- The Daily Update
- Posts
- Apple reaches deal with OpenAI
Apple reaches deal with OpenAI
AND OpenAI inks training agreement with Reddit
Jack Shields
May 21, 2024
Welcome back!
Big news today. Apple and OpenAI just reached an agreement that could bring us the world’s first (actually useful) personal AI assistant. That and more…
In today’s Daily Update:
🗞️ Apple and OpenAI plan major WWDC announcement
🤖 How generative AI is threatening the 2024 US Election
📸 OpenAI reaches AI training deal with Reddit
🚨 AI Roundup: Four quick hits
Read time: 2 minutes
TOP STORY
🗞️ Apple and OpenAI plan major WWDC announcement
WWDC = Apple’s annual developer conference in June (Image generated by DALL-E 3)
Apple and OpenAI have reportedly reached an official agreement to integrate OpenAI’s tech into iOS 18.
What you should know:
Apple plans to announce improvements to Siri, automatic summaries of notifications and news articles and transcriptions of voice memos.
The company hopes to launch “proactive intelligence” features that help users with their day-to-day lives.
OpenAI is expanding capacity to prepare for a massive influx of new users.
This agreement is not seen as a long-term solution, but Apple believes that its own AI models combined with OpenAI are sufficient for now.
Why it matters: This partnership is a win for both companies. Apple is on track to deliver proactive intelligence to its customers, and OpenAI will make its products available to almost 1.5 billion iPhone users around the world.
AI INSIGHT
🤖 How generative AI is threatening the 2024 US Election
Patrick Semansky | Associated Press
The U.S. Department of Homeland Security (DHS) just released a new analysis that outlines how AI could threaten the general election in November.
Key points:
DHS says generative AI tools enable domestic and foreign threat actors to interfere with the election “by aggravating emergent events, disrupting election processes or attacking election infrastructure.”
Attempts have already been made to conduct cyber attacks and spread disinformation online with AI.
Deepfaked images, videos and audio were highlighted as key threats to voters and election staff.
DHS Assistant Secretary Elizabeth Neumann urged voters not to believe anything seen on social media.
Why it matters: This election cycle is already getting messy after a robocall impersonating President Joe Biden circulated in New Hampshire in January. AI-driven efforts to mislead voters will continue to ramp up in the coming months, making it important to be cautious when you scroll social media and consume traditional media.
BUSINESS SPOTLIGHT
📸 OpenAI reaches AI training deal with Reddit
DALL-E 3
OpenAI has inked a deal with Reddit to use the social media platform’s content to train AI models.
The details:
Reddit will give OpenAI access to posts and replies in real time. This content will be incorporated into OpenAI’s ChatGPT.
OpenAI will become a Reddit advertising partner and help the platform launch new AI-powered features for users and moderators.
Reddit stock climbed more than 10% following news of the deal.
OpenAI CEO Sam Altman has a 8.7% stake in Reddit, but reportedly excluded himself from this decision.
Why it matters: Reddit is a gold mine for training data in the eyes of OpenAI. Social media platforms are rich sources of conversational and informal language that can help LLMs like ChatGPT generate more humanlike responses.
MORE TRENDING NEWS
🚨 AI Roundup: Four quick hits
Source: Microsoft
ChatGPT’s mobile app revenue spikes 22% after the launch of GPT-4o.
Microsoft introduces “Copilot+” PCs that boast advanced AI features.
OpenAI pauses ChatGPT’s Scarlett Johansson-like voice.
European Union threatens Microsoft with a billion-dollar fine over AI-generated misinformation in Bing.
THAT’S ALL FOR TODAY