- Horizon AI
- Posts
- Claude's New Model, Better than GPT-4o? 💥
Claude's New Model, Better than GPT-4o? 💥
+ PathChat 2: The Future of Medical AI
In partnership with
Welcome to another edition of Horizon AI,
Anthropic surprises everyone with a new model that appears to outperform GPT-4o.
Let's get into it!
Read Time: 3.0 min
Here's what's new today in the Horizon AI
A Chart to Kickstart Your Day 📈: Usage of AI Among Ecommerce Marketers Worldwide
Anthropic Launches Claude 3.5 Sonnet 💥
AI Revolution in Medical Diagnosis: PathChat 2
AI Tools to check out and AI Job board
The Latest in AI and Tech 💡
AI Showcase
A chart to kickstart your day
Usage of AI Among Ecommerce Marketers Worldwide
Data was obtained from the May 2024 Klaviyo and Qualtrics report, "Ecommerce Marketers' AI Trends."
1,510 e-commerce marketers from Australia, France, New Zealand, the UK, and the US were surveyed.
AI News
ANTHROPIC
Anthropic Launches Claude 3.5 Sonnet 💥
Anthropic has just launched Claude 3.5 Sonnet, a new model that promises unparalleled performance, speed, and cost-effectiveness. This new model sets industry standards and even outperforms GPT-4o in several benchmarks.
Details:
Claude 3.5 Sonnet is twice as fast as Claude 3 Opus, excelling in benchmarks for graduate-level reasoning (GPQA), undergraduate-level knowledge (MMLU), and coding skills (HumanEval).
In internal assessments, Claude 3.5 Sonnet successfully solved 64% of coding tasks, compared to 38% by Claude 3 Opus. It can write, edit, and execute code independently.
The model demonstrates significant improvements in understanding nuance, humor, and complex instructions, producing high-quality, natural-sounding content.
Claude 3.5 Sonnet also outperforms its predecessor in visual logic tasks, such as interpreting diagrams and transcribing text from imperfect images.
Claude 3.5 Sonnet is now accessible for free on Claude.ai and the Claude iOS app, with higher usage limits for Pro and Team subscribers. It is also available through the Anthropic API, Amazon Bedrock, and Google Cloud's Vertex AI.
TOGETHER WITH COURSERA
Boost Your Career with Coursera Plus
Looking to expand your skill set and stand out in your field? For a limited time, Coursera Plus is offering 40% off their subscription plan.
What's in it for you?
Over 7,000 courses from top companies like Google, Microsoft, and IBM
Learn skills that can lead to better pay and new job chances
Join the 77% of learners who say Coursera helped their careers
Don't wait - this offer expires on June 23, 2024. don't miss out. Click here to learn more and save 40% on Coursera Plus.
AI RESEARCH
AI Revolution in Medical Diagnosis: PathChat 2
The Mahmood Lab at Brigham and Women’s Hospital has introduced PathChat 2, a groundbreaking medical large language model that serves as a virtual consultant for pathologists. This innovative model aids in the identification, assessment, and diagnosis of tumors and other serious conditions.
Details:
PathChat 2 excels in multiple-choice diagnostic questions and generates clinically relevant responses to open-ended inquiries, outperforming leading models like ChatGPT-4V, LLaVA, and LLaVA-Med.
The model understands both pathology images and clinically relevant text, it can basically “have a conversation with a pathologist.”
Despite its breakthroughs, PathChat 2 still faces problems with hallucinations. Researchers suggest that reinforcement learning from human feedback and continuous training with up-to-date knowledge are necessary to mitigate this issue.
PathChat 2 is available through an exclusive license with Modella AI, a Boston-based biomedical AI company. The researchers note that it is still a work in progress, and they plan to improve the model and even expand its use cases.
AI Tools to check out
🌐 SiteForge: Easily generate sitemaps, wireframes and content for websites in minutes.
💒 Remodel AI: Remodel your home with AI
✅ Airtrain: Query and compare a large selection of open-source and proprietary models at once.
📊 Flowsage: Effortlessly design, collaborate, and optimize flowcharts with the power of AI
✨ SynthesysX: Create your own versions of any image you find online.
AI Job Board
💻 Software Engineer - Stripe
📈 Product Manager - Union
👨💻 Software Engineer - Bubble
💻 Product Manager - Deepgram
The latest in AI and Tech
Former Snap engineer Vu Tran has introduced Butterflies, a unique social network where humans and AIs engage through posts, comments, and DMs. The app launched on Tuesday on iOS and Android after a five-month beta. It allows users to create AI personas, called Butterflies, which autonomously generate content, fostering interaction between human and AI users.
Oracle announced a $1 billion investment over the next decade in AI and cloud computing in Spain. This initiative includes establishing a new cloud region to help customers transition workloads to Oracle Cloud Infrastructure and comply with EU regulations like DORA. This will be Oracle's third cloud region in Madrid, in partnership with Telefónica España.
OpenAI CTO Mira Murati predicts that AI systems could achieve PhD-level intelligence in certain fields within 18 months. She compares older models like GPT-3 to toddlers and describes the leap from GPT-4 to GPT-5 as akin to someone progressing from high school to university.
AI video startup HeyGen has secured $60 million in a Series A funding round led by Benchmark, elevating its valuation to over $500 million. HeyGen enables companies to create, localize, and personalize videos using virtual avatars without the need for a camera or crew, making video production more accessible and efficient.
AI Showcase
🖼 AI art of the day
Source: phac17 on Midejourney
In the style of Anthony Burrill, portrait of Maltese dog wearing sunglasses --ar 3:4
📹 AI video of the day
DREAM MACHINE
@LumaLabsAI's dream machine image to video is absolutely magical - aesthetically right up my alley!
→ img: midjourney v6 + --p
→ vid: Luma AI Dream Machine
→ song: Dawn of Silence by Kyon— Julie W. Design (@juliewdesign_)
4:56 PM • Jun 13, 2024
🧠 AI creation of the day
Book Covers by Leonardo Phoenix 📔
Leonardo Phoenix is capable of making stunning text for your images. I am excited to share some book covers I made with this new foundation model of @LeonardoAi_
The best part of it is the AI understanding of your simple prompt and it will… x.com/i/web/status/1…
— James Yeung (@jamesyeung18)
10:03 AM • Jun 18, 2024
That’s a wrap!
We'd love to hear your thoughts on today's email!Your feedback helps us improve our content |
Not subscribed yet? Sign up here and send it to a colleague or friend!
See you in our next edition!
Gina 👩🏻💻