- 80/20 AI
- Posts
- The Pentagon Deal That Broke AI
The Pentagon Deal That Broke AI
Gemini 3.1 Flash-Lite — Fast, Cheap & Ready to Deploy
Alamin Hossain
March 05, 2026
Advertise here | 6-min Read
Welcome to our Thursday edition!
In today’s menu:
AI Inspirational Quote
The Pentagon Deal That Broke AI
Gemini 3.1 Flash-Lite
Today's AI Radar
Daily AI Byte
Trending AI Tools
Social Media
"The measure of intelligence is the ability to change."
— Albert Einstein
The better question is who will build it without letting you make avoidable mistakes.
Experienced professionals building their first software venture often underestimate the product decisions made before development begins.
Before you commit capital, we pressure-test the idea properly.
If it survives, we build it right.
100+ founders trust Byldd.
🔥 Today's Biggest Story
Anthropic's CEO went on record accusing OpenAI of misrepresenting their Pentagon agreement. The tension has been building for weeks — and it just went public.
The core dispute: Anthropic refused to sign a DoD contract that lacked hard restrictions on autonomous weapons and civilian surveillance. OpenAI moved fast to fill the gap — then publicly claimed their deal includes those same protections. Amodei says that's not accurate.
Three reasons the public agrees with Anthropic:
The "any lawful purpose" clause in both contracts is a moving target — laws change, and yesterday's protection may not exist tomorrow
OpenAI's restrictions are tied to current law, not ethical red lines — a crucial difference when governments rewrite rules fast
Users voted with their phones: ChatGPT saw a 295% spike in uninstalls within 48 hours of the deal going public
Claude is now the #1 downloaded app in the US. The market has spoken — people want to know their AI tools have a spine. For business leaders, this is no longer just a tech story. It's a vendor risk and brand alignment question.
Bottom line : This isn't just an ethics debate — it's a business risk conversation. Every company using AI tools right now needs to ask one question: if my AI vendor's contract became front-page news tomorrow, would I be comfortable with what it says? Anthropic just made that question unavoidable.
Google quietly dropped a model this week that enterprise teams and developers have been waiting for — something powerful enough to matter, cheap enough to actually deploy at scale.
Why this one's different:
Priced at just $0.25 per million input tokens — the most affordable model in the entire Gemini 3 lineup
Runs 2.5x faster than Gemini 2.5 Flash and generates output 45% quicker — critical for high-frequency, real-time workflows
Beats GPT-5 mini and Claude 4.5 Haiku on 6 out of 11 benchmarks, including graduate-level science questions
Handles text, images, speech, and video in a 1 million-token context window — built for translation, content moderation, UI generation, and simulations
Available now in preview via Google AI Studio and Vertex AI. Not for complex agent reasoning — but for everything you need done at volume, fast, and at a price that actually makes sense to deploy.
Bottom line: Most teams are still paying premium prices to run AI on tasks that don't need it. Gemini 3.1 Flash-Lite changes that calculation completely. If your team processes content, moderates data, or builds real-time pipelines at scale — this is the model to test this week. The price-to-performance gap just got very hard to ignore.
The Hustle: Claude Hacks For Marketers
Some people use Claude to write emails. Others use it to basically run their entire business while they play Wordle.
This isn't just ChatGPT's cooler cousin. It's the AI that's quietly revolutionizing how smart people work – writing entire business plans, planning marketing campaigns, and basically becoming the intern you never have to pay.
The Hustle's new guide shows you exactly how the AI-literate are leaving everyone else behind. Subscribe for instant access.
📡 Today's AI Radar
Anthropic's $60B investor base is now at risk — here's what that means
Over 200 venture capital firms poured $60 billion into Anthropic. That entire investment is now under threat because of the Pentagon blacklist designation. If enforced fully, it would bar NVIDIA, Amazon, and other military contractors from doing business with Anthropic — potentially cutting off the cloud infrastructure the company runs on.
Scientists built AI that mimics human personality — with 98.7% accuracy
Researchers developed PsychAdapter, a system that fine-tunes large language models to reflect specific personality traits and mental health profiles.. The line between AI behaviour and human psychology just got a lot thinner.
Andrew Ng: AGI is decades away — and the real bubble risk is in training infrastructure
AI legend Andrew Ng pushed back on AGI hype this week, stating that AI capable of matching the full breadth of human intelligence remains far off. . A rare voice of restraint in a very loud room.
Samsung's Galaxy AI goes deeper at MWC 2026 — 800M devices by year-end
Samsung showcased its Galaxy S26 series at MWC 2026 this week, highlighting how Galaxy AI now works across phones and wearables to anticipate user needs before they ask. The company is targeting 800 million AI-enabled devices by end of 2026 in partnership with Google.
🧠 Daily AI Byte
🔧 Thursday Skill — Tool: Claude
Your first draft shouldn't take 2 hours. Make Claude write it in 2 minutes.
Most leaders spend the bulk of their writing time staring at a blank screen — not actually writing. The fix isn't to write faster. It's to never start from zero. Give Claude a rough brain dump — messy notes, bullet points, even a voice memo transcript — and ask it to turn that into a sharp first draft. You edit, refine, and add your voice. Claude does the structural heavy lifting. The result: better output in a fraction of the time, and your thinking stays front and centre.
"Here are my rough notes: [paste your messy thoughts]. Turn this into a clear, direct 200-word update for my leadership team. Keep my voice — no corporate fluff, no filler sentences."/
Social media
🔧 Trending AI Tools
Kodo — Forget drag-and-drop. Kodo lets you build fully editable designs just by describing what you want in a chat.
AssemblyAI — If you're building any kind of voice agent or real-time audio product, AssemblyAI just raised the bar.
Locally AI + Qwen — Alibaba's latest Qwen models, running directly on your iPhone. No cloud. No API costs.
The Ultimate ChatGPT Prompt Collection — A free, structured library built to help anyone get dramatically better results from AI — faster.
Vocova — Pulls accurate transcripts from audio and video across 1,000+ platforms in one place. Podcasts, calls, meetings, YouTube — all searchable, all documented.
👋 THAT’S A WRAP
SPONSOR US
Get your business in front of over 90k+ AI professionals
8020AI is the world’s #1 AI Newsletter, Read by 90k+ professionals from leading companies such as Google, OpenAI, Meta, and Microsoft.
We've assisted in promoting Over 500 AI-Related Products. Will yours be the next?
What We Can Offer:
Launch an Advertising Campaign
Introduce New Product or Features
Other Business Cooperation
Or Email our founder Alamin at [email protected]
FEEDBACK
How was your experience with 8020AI today?
How was 8020AI today? |
If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.
