- #TheAlpha Newsletter
- Posts
- The DeepSeek Dilemma: The U.S. is Freaking Out Over AI—But Are We Any Better?
The DeepSeek Dilemma: The U.S. is Freaking Out Over AI—But Are We Any Better?
Plus, upcoming creator economy events to check out!
DeepSeek, a Chinese AI startup, has recently made waves with its R1 model, which rivals leading AI systems like OpenAI's ChatGPT. However, this achievement has now sparked controversy, especially with its US competitors.
OpenAI and Microsoft are investigating whether DeepSeek improperly used data from OpenAI to develop its AI models. They suspect DeepSeek employed a technique called "distillation," where a smaller model learns from a larger one, potentially violating OpenAI's terms of service.
Critics argue that DeepSeek's data collection practices, which involve storing user data on servers in China, pose severe privacy risks. This has led to concerns about the potential misuse of collected data by the Chinese government.
Interestingly, OpenAI has faced similar criticisms related to stealing content. Its models have been trained on vast amounts of internet data and collect just as much, raising questions about data privacy and consent. This situation is another obstacle, especially with media publishers, creators, and the entertainment industry at large.
So is DeepSeek really the villain here, or is this just a case of the pot calling the kettle black—where all these AI models have the same issues but seem more innocent simply because they’re American?
deepseek's r1 is an impressive model, particularly around what they're able to deliver for the price.
we will obviously deliver much better models and also it's legit invigorating to have a new competitor! we will pull up some releases.
— Sam Altman (@sama)
2:29 AM • Jan 28, 2025
For context, here are some differences in terms of data collection and storage between OpenAI and Deepseek:
DeepSeek stores all user data on servers in China, while OpenAI's storage locations are not explicitly detailed.
DeepSeek may share user data with its corporate group and comply with legal obligations, potentially involving data disclosure to Chinese authorities, whereas OpenAI does not use data submitted through its API to train or improve models unless users opt-in.
DeepSeek's data storage in China raises concerns due to local laws that may require companies to share data with government authorities upon request, while OpenAI implements data encryption both at rest and in transit, adhering to industry-standard security practices.
Congress is already taking steps towards continuing to stop this possible AI threat from China, with Republican Senator Josh Hawley introducing the Decoupling America’s Artificial Intelligence Capabilities from China Act, which would:
“ Prohibit the import from or export to China of artificial intelligence technology; prohibit American companies from conducting AI research in China or in cooperation with Chinese companies; and prohibit U.S. companies from investing money in Chinese AI development.”
DeepSeek also turned heads by claiming they developed their DeepSeek-V3 model for just $6 million—a fraction of what similar projects cost in the West. They managed this by using the more affordable Nvidia H800 GPUs, tailored for the Chinese market due to export restrictions. This has sparked lively discussions and a bit of skepticism among tech enthusiasts, wondering if such a breakthrough could really come at such a low price.
🔥My Hot Take🔥: We've seen this pattern with electric cars, social media platforms like TikTok, and now AI. The U.S. will continue finding ways to cut China out, fueling an ongoing debate over access, innovation, and national security—often at odds with public sentiment. While concerns about AI from China are valid, the bigger issue may be the public’s understanding of why these restrictions are happening. At the same time, focusing solely on China can be a distraction from similar concerns with companies based in the U.S., particularly around data privacy and misuse—issues we shouldn’t sweep under the rug.
Other headlines to check out:
AI
LinkedIn co-founder Reid Hoffman just raised $25 million to take on cancer with AI - QZ
LinkedIn lawsuit over use of customer data for AI models is dismissed - Reuters
AI: Europe Imposes Its Framework Starting This Sunday - CoinTribune
California is investigating OpenAI’s conversion to a for-profit company - CalMatters
Creator Economy
Instagram tests Reels pause feature as TikTok remains in limbo - CNBC
How Small Businesses Can Adapt to TikTok’s Uncertain Future - Inc
Apple reportedly gives up on its AR video glasses project - The Verge
Reese Witherspoon Brings Hollywood Glam to Networking With Career Update - Parade
DeepSeek Is Great For Creators - Jim Louderback
Instagram boss debunks popular ‘hack’ for boosting likes and views - DailyMail
Web3
Trump finds a new way for foreign governments to pay him off: Crypto - The Washington Post
It's Easier Than You Think to Build With AI and Web3 - CoinDesk
ai16z’s Eliza Labs ships white paper for Web3-native AI agents - Cointelegraph
Keeping my eye on… Upcoming events:
SoCom - Feb 6

SXSW March 7-15
Plus, don’t miss my panel, “Creators for Mental Health: Building Wellness in the Digital Age,” happening on Tuesday, March 11, at 4 PM!
Gentle Reminder 🙏
When the news gets overwhelming, it's okay to take a break and look after yourself. Disconnecting from the constant updates gives you a chance to recharge and keep your peace of mind. Remember, it’s not just okay to take some time for yourself—it’s necessary.
Remember, I'm Bullish on you!
With gratitude,
