AI Running Buzzfeed

PLUS: Anthropic’s Claude AI Can Now Digest An Entire Book In Seconds

Good morning! Today we're covering BuzzFeed's AI-generated quizzes outperforming human-created ones, advancements for Anthropic's AI, Claude, that can now analyse books in seconds and experts claiming autonomous vehicles are risky due to a lack of evidence of accident mitigation.

Let's get to it!

Top News & Why It Matters

| BuzzFeed Says Its AI Is Outperforming Its Pesky Human Employees

BuzzFeed claims its readers spend 40% more time on AI-generated quizzes compared to traditional ones, according to Bloomberg. The media company is focusing on AI after shutting down its Pulitzer-winning news division and laying off around 120 employees.

BuzzFeed has been experimenting with AI to "enhance human creativity" and generate content, including quizzes and SEO-driven travel guides. However, the AI-generated content has yet to live up to the "more personalized, more creative, more dynamic" promise made by CEO Jonah Peretti. futurism

Why does this matter?

While BuzzFeed's experimentation with AI-generated content may attract users, it remains to be seen if it can maintain journalistic quality and avoid errors or plagiarism in AI-generated articles.

| Anthropic’s Claude AI Can Now Digest An Entire Book Like The Great Gatsby In Seconds

AI company Anthropic announced that its AI language model, Claude, can now analyze an entire book in under a minute thanks to an expanded context window. Similar to OpenAI's GPT-4, Claude is a large language model that predicts the next token in a sequence given a certain input.

Increasing Claude's context window to 100,000 tokens, or about 75,000 words, has allowed it to process larger works like books and participate in lengthy interactive conversations. Anthropic claims that Claude can now engage in conversations spanning hours or even days. arstechnica

Why does this matter?

This rapid analytical capability could revolutionize how we interact with AI, leading to more comprehensive and nuanced conversations as well as the possibility of AI becoming more involved in literary and research fields.

| A Former Pilot On Why Autonomous Vehicles Are So Risky

Missy Cummings, an expert in transportation safety, recently spoke about the risks and realities of advanced driver-assistance systems (ADAS) in cars. She pointed out that there is no evidence of accident mitigation with ADAS-equipped cars and that people using these systems are more likely to be speeding if an accident occurs.

Cummings also discussed the concept of "risk homeostasis," where people believe the car will do certain functions for them, only to find out it doesn't. This results in drivers taking more risks because they assume the technology will keep them safe. spectrum

Why does this matter?

Although advanced driver-assistance systems have the potential to improve road safety, their usage might make drivers complacent or too reliant on them. Policymakers and car manufacturers must clearly define the limitations and capabilities of ADAS to ensure that drivers remain alert and responsible while on the road.

| Together With WriteWell

WriteWell is your fastest route to becoming a world-class communicator. It’s a custom AI trained on thousands of high-quality writing samples that goes beyond the capabilities of ChatGPT and other writing apps to make your writing concise.

Get 50 free credits when you sign up now using this link.

Mastering Artificial Intelligence

| Our Top Picks

✅ Watermelon: Transform customer service with AI and chatbot tech, GPT-4 integration. (website)

✅ Noise Eraser: AI-powered video noise remover with adjustable audio volume. (website)

✅ Digital Roughnecks Podcast: Energy podcasts with expert guests; schedule your own episode. (website)

✅ Certainly: Digital twin for personalized conversations improving conversion rates and reducing returns. (website)

| Lunch & Learn

📝 Text: Machine learning, explained by Sara Brown, MIT (link)

🎧 Audio: [AI Strategy] Approaching Strategy Top-Down and Bottom-Up - with Bret Greenstein, Partner at PwC (link)

📽️ Video: DeepLearning.ai (link)

What Else Is Happening

| More News

🗞️ Google’s A.I. Bonanza and Driverless Car Talk With Cruise C.E.O. Kyle Vogt. (source)

🗞️ Allen Institute raises $30m to invest in AI-focused start-ups. (source)

🗞️ Stability AI releases an open source text-to-animation tool. (source)

🗞️ AI Is Speeding Up Astronomical Discoveries. (source)

🗞️ You’re Probably Underestimating AI Chatbots. (source)

🗞️ Google Bard hits over 180 countries and territories—none are in the EU. (source)

🗞️ Google’s new Magic Editor pushes us toward AI-perfected fakery. (source)

🗞️ Google Bard AI now generally available. (source)

| Best of AI Social Media

💬 AI celebrates Mothers as irreplaceable on Mother's Day (Twitter)

💬 "Bard's Massive Update Surpasses ChatGPT: 8 New Features" (Twitter)

💬 1930s Lower Manhattan Waterfront footage restored (Twitter)

💬 Asked for ASCII art "sunset" wtf (Reddit)

💬 Who are you? (Reddit)

💬 I have 15 years of experience and developing a ChatGPT plugin is blowing my mind (Reddit)

💬 Sundar Pichai's response to "If AI rules the world, what will WE do?" (Reddit)

💬 I fed GPTZero a letter from 1963 that went viral a few days ago and it marked it as AI written. (Reddit)

| Research Review

Design, Development, And Evaluation Of An Interactive Personalized Social Robot To Monitor And Coach Post-Stroke Rehabilitation Exercises (Full Paper)

Min Hun Lee et al.

∑ Summary: This paper presents a social robot exercise coaching system for personalized rehabilitation that uses a neural network and rule-based model and improves its performance by tuning with patient data.

!? Practical implications: The development and implementation of social robot exercise coaching systems for personalized rehabilitation in post-stroke survivors has the potential to lessen healthcare costs and workloads, enabling medical professionals to address a larger number of patients more efficiently.

The Asnr-Miccai Brain Tumor Segmentation (BraTS) Challenge 2023: Intracranial Meningioma (Full Paper)

Dominic LaBella et al.

∑ Summary: The BraTS meningioma 2023 challenge aims to develop automated intracranial meningioma segmentation models using the largest annotated mpMRI dataset, ultimately improving patient care.

!? Practical implications: This research could potentially streamline the diagnostic process for meningiomas by providing automated MRI segmentation tools, leading to faster and more accurate diagnoses and ultimately improved patient outcomes in adults with intracranial tumours.

Written by Isaac R. Ward, Casey Clifton, and Alex Brogan.

Send us your feedback at [email protected]. We’re ready to listen.

Join the conversation

or to participate.