- AI Fire
- Posts
- 🎭 AI Can Read Your Emotions
In partnership with
Plus: The State of AI Report by McKinsey
Read time: 5 minutes
China is developing a quantum chip they claim is 1 quadrillion times faster than the world’s top supercomputers, including Google’s. Has the era of quantum computing already begun?
What are on FIRE 🔥
AI INSIGHTS
🤖 The State of AI: How Organizations are Rewiring to Capture Value
Companies are starting to do things that help make more money—like changing how work gets done with gen AI and putting top bosses in charge of AI rules.
Key takeaways:
Workflow Changes Boost Profits, but Impact Is Still Small - Redesigning workflows is key to gen AI profits (EBIT), with 21% of users making big changes. Yet, over 80% see no company-wide impact, showing gen AI is still early.
CEOs Drive Success at Big Firms - CEO oversight of AI governance lifts EBIT, especially at firms with ≥$500M revenue (28% report CEO involvement, 17% board oversight). Larger companies adapt faster.
AI Use Increases Across Businesses - Now 78% of organizations use AI in at least one area (up from 55% in 2023), with gen AI at 71% (up from 65% in 2024). It’s in three functions on average, topping out in IT (36%), marketing/sales (42% for gen AI), and service operations.
Risks and Skills Get More Attention - Firms are tackling gen AI risks like errors, hacks, and IP issues more than in early 2024, led by big companies. 13% hired AI compliance experts, reskilling is up, but 50% foresee a data scientist shortage next year.
Big Firms Lead with Smart Strategies - Companies with ≥$500M revenue excel at best practices (e.g., KPIs tracked by <20%, roadmaps by <33%), hire more talent, and cut staff to raise EBIT. Smaller firms trail behind.
Why it matters: Big companies are ahead because they’ve got the cash, people, and plans to make it work. Smaller ones are slower, which could leave them in the dust if they don’t catch up. Workflow changes and boss involvement are the secret sauce for profits, but it’s still early—2025 might be when things really pop. The talent crunch and risk stuff? That’s the tricky part everyone’s got to figure out soon.
PRESENTED BY ARTISAN
Try Artisan’s All-in-one Outbound Sales Platform & AI BDR
Ava automates your entire outbound demand generation so you can get leads delivered to your inbox on autopilot. She operates within the Artisan platform, which consolidates every tool you need for outbound:
300M+ High-Quality B2B Prospects, including E-Commerce and Local Business Leads
Automated Lead Enrichment With 10+ Data Sources
Full Email Deliverability Management
Multi-Channel Outreach Across Email & LinkedIn
Human-Level Personalization
TODAY IN AI
AI HIGHLIGHTS
✨ Manus AI can be your personal travel planner – This chatbot crafts a 3-day Hanoi trip under 15M VND (about $590), using human-like skills! It thinks (plans, tracks, recovers from errors), searches across multiple engines, and browses websites to extract info, book, and suggest activities. Whether your budget is modest or big, it ensures you have the best trip possible. You can view its process here.
📈 Anthropic is doing well this year with strong revenue, thanks to its role in powering the popular new app, Manus. Their business is growing fast as a result.
🦾 Despite having 21 billion parameters, which is smaller than many other models, Reka Flash 3 performs exceptionally well, especially for tasks like general chat, coding, and following instructions. It is competitive with larger proprietary models like OpenAI’s o1-mini, making it a great choice for low-latency applications.
🤖 Sam Altman shared that OpenAI has trained a new model that's really good at creative writing. He even said, "It's the first time I’ve been truly impressed by something written by AI."
⚡ Nous Research, a New York-based AI collective, launched an Inference API for easier access to its models. Unlike big tech companies, they focus on unrestricted, personalized, and transparent AI, giving users more control. Join their waitlist here.
🎭 Alibaba just launched a new AI model called R1-Omni. It can read emotions by analyzing a person’s facial expressions in videos, adding emotional intelligence to computer vision. And the best part? You can download it for free on Hugging Face.
⚛️ Researchers at the University of Science and Technology of China (USTC) introduced the Zuchongzhi-3, a superconducting quantum computing prototype. This quantum chip is 1 quadrillion times faster than the world’s fastest supercomputers and 1 million times faster than Google’s Sycamore processor.
💰 Daily AI Fundraising: Insilico Medicine raised $110 million in Series E funding, bringing its valuation to over $1 billion. Led by Value Partners Group, the funds will support AI-designed drugs, including a promising IPF candidate.
AI SOURCES FROM AI FIRE
NEW EMPOWERED AI TOOLS
🖼️ Adobe’s Customize feature allows you to tweak stock images, or even multiple images, to fit your needs.
🎙️ Sonic 2.0 can capture the full details of your voice, including accents, with just a few seconds of audio.
📸 Inductive Moment Matching from Luma Labs is a new technique that generates superior-quality images way faster than before.
🤖 Maestro by AI21 Labs plans and handles complex tasks with precision, control, and reliability.
💼 Harvey brings agentic workflows that work with professionals to deliver accurate, purpose-driven results.
AI QUICK HITS
🏋️ Google’s AI Brings Robots to Life with Human-Like Dexterity and Smarter Decision-Making (Link)
🤖 ServiceNow Launches No-Code, Low-Code AI Agent Builder (Link)
🎥 Snapchat Introduces AI Video Lenses for Platinum Users (Link)
🏀 DeepMind’s AI Robot Nails a Perfect Slam Dunk (Link)
💻 Anthropic CEO: AI to Write 90% of Code Soon (Link)
AI CHART
Early 2025 AI Ecosystem - OpenAI and Anthropic Dominate, but Competition Grows
Market Leaders: OpenAI and Anthropic are the big dogs in text on Poe, grabbing about 85% of all messages. They’re killing it with super chatty AI!
Anthropic’s Jump: Anthropic’s stuff really took off in 2024, especially after dropping Claude 3.5 Sonnet in June. Now it’s neck-and-neck with OpenAI for text messages—competition’s heating up!
Model Cannibalization: When a new top model comes out, people ditch the old one fast. Like, GPT-4 got swapped for GPT-4o, and Claude-3 lost out to Claude 3.5. Everyone just wants the shiny new thing!
DeepSeek’s Rise: These open-source DeepSeek models, R1 and V3, went from nothing in December 2024 to 7% of messages at their peak. That’s way more than old-timers like Llama or Mistral ever got!
Google and Others: Google’s Gemini stuff was picking up messages until October 2024, but then it started slipping. Meanwhile, a bunch of smaller models are popping up, giving users more fun options to play with.
AI CHEAT SHEET
AI JOBS
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