- AI Fire
- Posts
- 💊 Unlock $2.75B AI Pharma Bet
In partnership with
Plus: 5 Practical Claude Dispatch Use Cases That Run While You're Away From Your Desk (Steal Them) - Preconfigured Versions, Ready to Use
A new Stanford study warns that your favorite chatbot (GPT, Gemini or even Claude) might be making you a bit too full of yourself by agreeing with everything you say.
What's on FIRE 🔥
IN PARTNERSHIP WITH SECTION
Join Section on April 23 for an afternoon of free workshops that will help you become part of a class of leaders who can rise to the challenge.
You'll get frameworks for driving successful AI adoption, advice from leaders who have done this before and a community of ambitious Heads of AI.
AI INSIGHTS
Eli Lilly just signed a $2.75B partnership with Insilico Medicine to commercialize drugs discovered using generative AI. Insilico has already produced 28 AI-generated drug candidates, with many reaching clinical-stage validation.
Pharma companies are moving AI from a “nice-to-have tool” to a core layer of the drug discovery stack. The structure of the deal also tells an interesting story:
Deal size: up to $2.75B total value
Upfront payment: $115M, with milestone-based payouts later
Pipeline: 28 AI-designed drug candidates
Stage: nearly half already in clinical development
Partnership history: Lilly and Insilico have collaborated since 2023
Structure: royalties + milestone payments tied to regulatory success
This structure usually means both sides see strong potential, but want risk-sharing as trials progress. Insilico focuses on generative biology models. Lilly brings global development infrastructure.
Insilico’s CEO even noted Lilly has strong internal AI capabilities combining multiple scientific domains into unified systems. AI is starting to influence what medicines get invented in the first place.
PRESENTED BY WISPR FLOW
Speak your prompts. Get better outputs.
The best AI outputs come from detailed prompts. But typing long, context-rich prompts is slow - so most people don't bother.
Wispr Flow turns your voice into clean, ready-to-paste text. Speak naturally into ChatGPT, Claude, Cursor, or any AI tool and get polished output without editing. Describe edge cases, explain context, walk through your thinking - all at the speed you talk.
Millions of people use Flow to give AI tools 10x more context in half the time. 89% of messages sent with zero edits.
Works system-wide on Mac, Windows, iPhone, and now Android (free and unlimited on Android during launch).
AI SOURCES FROM AI FIRE
1. Free: How to Build Hyperspecific AI Apps Without Expensive Subscriptions (Save $3,600 a Year). Move beyond ChatGPT. Discover how to build and host custom, hyperspecific AI agents & apps using AI. No-code friendly. 100% yours.
2. Free: New FREE Google Gemini 3.1 Flash Can Talk & See Everything At Once Without Any Lag. Google Gemini 3.1 Flash is very fast. I used it to talk and share my screen without waiting. This page shows you how to use it for free today now.
3. 5 Practical Claude Dispatch Use Cases That Run While You're Away From Your Desk (Steal Them). How to trigger lead scraping, meeting prep, inbox sweeps, and research directly from your phone using Claude Dispatch.
HOW TO RUN OPEN-SOURCE AI MODELS
Yes, every time you send a message to ChatGPT or Claude, your data passes through someone else's server. You also pay for every single word the AI writes back.
In 2026, open-source AI has reached a tipping point. You can run models close to GPT-level quality without paying per prompt and without sending your data to external providers.
If you've been meaning to try open source AI models but assumed you needed fancy hardware, this is the best explainer I've seen, simply the best "how to use open models" guide out there. You’ll learn:
why many now start with hybrid AI setups instead of only local or cloud
how open models are closing the gap with proprietary AI
tradeoffs nobody explains when they say “open-source AI is cheaper”
a simple decision framework to avoid wasting weeks testing the wrong stack
If you’ve ever wondered whether open-source AI is finally practical for normal users, this will change how you think about the space.
TODAY IN AI
AI HIGHLIGHTS
🎯 Why recruiters are making interviews ‘AI-free zones’? Just a note for you, they're pushing interviews back to human-only formats as AI CVs flood hiring pipelines.
🚀 OpenAI is building a desktop 'Superapp' combining ChatGPT, Codex, and its AI browser. One app to write code, run tasks, and automate workflows. Read more.
💰 ChatGPT ads already hit $100M annualized revenue in just 6 weeks after launch. However, most users still don’t see ads yet. So what if rollout expands globally?
🕳️ Many insiders warn Anthropic’s unreleased ‘Mythos’ could outperform current defenses. A Dark Reading poll ranks agentic AI as the #1 threat vector for 2026.
🎬 Google’s internal coding agent “Agent Smith” became so popular that access had to be restricted. It runs tasks autonomously. Agent Smith entered the Googleplex.
🏛️ Anthropic just scored an early court win after being blacklisted by the US government over Claude. Judges cited possible First Amendment retaliation.
💰 Big AI Fundraising: Harvey AI, a legal tech company, secured $200M in funding led by Singapore's GIC and Sequoia Capital, underscoring the growing interest in AI-driven legal solutions.
Forget switching tabs. Forget slow document workflows. Now you can prepare, send, and track eSignatures instantly directly inside your AI conversation.
With SignNow’s MCP integration, your AI can handle contracts, approvals, and forms end-to-end without leaving the chat.
No complex setup. No extra tools. No delays. Just upload your document… and send a legally binding signature request immediately.
Perfect for NDAs, client agreements, onboarding forms, and approvals - done in minutes, not days.
NEW EMPOWERED AI TOOLS
🧩 Clico is a powerful Browser extension designed to be your ultimate writing and productivity partner. It lives directly inside your browser.
📊 Sheet Ninja turns any Google Sheet into a live API in seconds. Your data already exists. Paste a link, get endpoints.
🎧 SUN turns any topic into personalized podcasts, audiobooks, or audio courses you can listen to while living your life.
🤖 Parallel Code runs 10 AI coding agents at once. Works with Claude Code, Codex CLI, Gemini CLI, or any CLI-based agent. Free, MIT licensed.
AI BREAKTHROUGH
We all know chatbots often agree with users too much, which is called ‘AI sycophancy’. A new Stanford study suggests it may reduce critical thinking and increase emotional reliance on AI advice.
Across scenarios involving relationships, ethical dilemmas, and social conflicts, AI responses validated user positions about 49% more often than humans would:
Users trusted agreeable AI responses more, even when advice was questionable
Participants became more confident in their own opinions, even in situations where they might be wrong
Exposure to validating responses made people less likely to apologize or reconsider decisions
Around 12% of U.S. teens already use AI for emotional support
Why does this happen? Because models are optimized to be helpful and engaging. People rated sycophantic models as more trustworthy and said they were more likely to return for future advice.
In practice, AI works best as a thinking partner. As models become more conversational and personalized, this “agreeability bias” will likely become a major topic in AI safety and alignment discussions.
We read your emails, comments, and poll replies daily
How would you rate today’s newsletter?Your feedback helps us create the best newsletter possible |
Hit reply and say Hello – we'd love to hear from you!
Like what you're reading? Forward it to friends, and they can sign up here.
Cheers,
The AI Fire Team
Reply