- OpenTools' Newsletter
- Posts
- ⚠️ Pentagon Adopts Banned AI
⚠️ Pentagon Adopts Banned AI
Plus Apple chooses Google and Amazon records everything
Reading time: 5 minutes
🗞️In this edition
Pentagon Adopts Musk's Grok Days After Global Deepfake Scandal
Apple Picks Google's Gemini to Power New Siri
Amazon Bets on Always-Listening AI Wearable
In other AI news –
AI Drug Startup Raises $25M From Meta and OpenAI Execs
Microsoft Teams With PayPal and Stripe for AI Shopping
Nvidia and Lilly Invest $1B in Drug Discovery Lab
4 must-try AI tools
Three decisions this week that don't make sense at first.
One organization is adopting technology that other countries just banned. A tech giant picked its biggest competitor over its closest partner. And a major company is showcasing a device that raises questions nobody seems to be asking.
Each story involves risk. Each involves speed. And each reveals something about who's really in control.
The pattern that emerges isn't what you'd expect. The organizations moving fastest aren't the ones you'd think would take the biggest chances.
This edition looks at decisions that reveal more than the people making them might want you to know.
What's happening:
Defense Secretary Pete Hegseth announced Monday that Grok will join Google's AI inside the Pentagon network. The chatbot will run on both classified and unclassified military systems.
The announcement came days after Grok drew global outcry for generating sexualized deepfake images without consent. Malaysia and Indonesia blocked the app, and the UK launched an investigation.
Hegseth said the Pentagon's AI will operate "without ideological constraints" and "will not be woke." He made the announcement at Musk's SpaceX facility.
The Pentagon will feed two decades of combat data into the system. The Pentagon has not responded to questions about Grok's deepfake and antisemitism issues.
Why this is important:
This shows the Pentagon prioritizing speed over safety. Grok generated explicit deepfakes days before getting access to classified military networks.
The timing is remarkable. Countries are banning Grok while the US military gives it access to combat intelligence.
Hegseth's "not woke" framing suggests choosing AI based on political positioning rather than technical capabilities. That's dangerous for military systems.
The Pentagon is trusting Musk's private company with two decades of combat data. No military accountability, enormous conflicts of interest.
Comments from the editor:
Hegseth announced Grok's Pentagon deployment without mentioning safety testing or evaluation. It's just going straight onto classified networks.
The "not woke" justification is alarming. Military AI should be evaluated on accuracy and security, not political ideology.
Musk's conflicts are massive. He owns X, runs SpaceX, controls Grok, and has enormous Pentagon contracts. Now his AI gets trained on military data.
Countries are blocking Grok while the US military embraces it. Either those countries see risks the Pentagon is ignoring, or the Pentagon is accepting risks they won't.
What's happening:
Apple and Google announced Monday a multi-year partnership to power Siri with Google's Gemini AI. The deal represents Apple's first time relying on another company to power core AI features.
Apple will pay roughly $1 billion per year for access to Gemini models, according to Bloomberg. The new Siri launches later this year.
The partnership comes after Apple tested OpenAI and Anthropic. Apple determined Google's technology provides the "most capable foundation" for its needs.
Apple already partners with OpenAI for complex queries. That deal isn't changing, but Gemini will power the core Siri experience.
The deal is not exclusive. Apple maintains all processing happens in Apple-controlled environments to preserve privacy.
Why this is important:
This shows Apple is behind in AI. Apple rarely uses competitors' technology for important features, especially not for Siri.
Google gets access to 1.5 billion iPhone users. That's huge distribution for Gemini beyond just search.
Apple promised a better Siri in 2026 but can't deliver it alone. This deal buys time while Apple builds its own AI for 2027.
This is bad news for OpenAI. Apple picked Gemini over ChatGPT as the main technology. That says something about which AI Apple thinks is better.
The $1 billion deal makes Apple and Google closer partners. Google already pays Apple billions to be iPhone's default search engine.
What's happening:
Amazon acquired Bee last July and showcased it at CES this week. The $50 wearable records conversations all day and now drafts emails and manages calendars.
Bee clips to your shirt or wears as a wristband. It records conversations, transcribes them in real time, then discards the audio. Users pay $19 monthly for AI features.
The updated version creates calendar invites and drafts emails based on what you said. If you mention following up with someone, Bee writes the email. If you discuss a meeting, it adds the calendar entry.
Amazon bought Bee to compete in wearables after previous failures. Alexa earbuds couldn't compete with AirPods and Ray-Ban Meta glasses. Bee extends Amazon's AI outside the home.
The eight-person Bee team joined Amazon but continues operating from San Francisco. Amazon says it will eventually merge Bee with Alexa, though timing is unclear.
Why this is important:
This shows Amazon's strategy for wearables is different from its smart speaker approach. Alexa sits in your home. Bee follows you everywhere.
The $50 price point challenges premium competitors. Humane's AI Pin cost $499 and flopped. Bee is cheap enough that curious buyers can experiment without major commitment.
Amazon gets access to extremely personal data. Conversations, emails, calendars, and daily patterns. That's more intimate than what Alexa collects sitting in your kitchen.
The always-listening aspect raises legal questions. Recording consent laws vary by state and country. Bee says audio isn't stored, but transcripts capture private conversations.
For Amazon, wearables represent the battleground against Apple, Meta, and Google. The CES showcase signals they're serious about making Bee a real competitor.
AI Drug Startup Raises $25M From Meta and OpenAI Execs – A new AI drug discovery startup just raised serious money from tech executives, joining what's becoming a surprisingly crowded race.
Microsoft Teams With PayPal and Stripe for AI Shopping – Microsoft's AI can now buy things for you, but there's a catch that's worrying online stores.
Nvidia and Lilly Invest $1B in Drug Discovery Lab – The pharmaceutical giant and chip maker are combining forces on something bigger than either could build alone, and it involves robots.
ThreadBois - An online tool that helps users create viral thread headers
Teach-O-Matic - A tool that helps users create how-to videos from text instructions
Watermelon - An all-in-one automated customer service tool powered by GPT-4
Audioatlas - Allows users to find the most suitable music from a vast global database of over 200M songs
The most interesting part isn't what's happening. It's what's not being said.
Why adopt something others are blocking? Why choose a competitor when you have options? Why showcase technology that creates obvious concerns?
The answers aren't in the press releases. They're in the timing, the contradictions, and what gets left out of the announcements.
These decisions create paths that can't be easily reversed. Once made, they change what's possible next.
If you have any feedback for us, please reply and let us know how we did. We're always looking to improve.
How did we like this version? |
Interested in featuring your services with us? Email us at [email protected] |