• Microsoft “doubling down” on cybersecurity “…engineering leads at Microsoft are now prioritizing security over new features or shipping products more quickly…”
• YouTube Q1 Ad Revenue Climbs 21% to $8.1 Billion
• Microsoft Q1 Cloud revenue $35.1 billion
• AI models inch closer to hacking on their own
• Apple releases eight small AI language models aimed at on-device use…Apple’s new AI models, collectively named OpenELM for “Open-source Efficient Language Models,”…
• JetBrains Launches IDE Services to Simplify Managing Development Tools
• Sam Altman, Jensen Huang, and the 20 other leaders on the new AI safety board
• Understanding API Technologies: A Comparative Analysis of REST, GraphQL, and Asynchronous APIs
• QCon London: Meta Used Monolithic Architecture to Ship Threads in Only Five Months
• Android TVs Can Expose User Email Inboxes
• DragonCrawl: Generative AI for High-Quality Mobile Testing
• Apple Reportedly Developing Its Own Custom Silicon for AI Servers

• Google consolidates its DeepMind and Research teams…Google is combining its Android and hardware teams — and it’s all about AI
• Netflix subscribers : 269.6 mio
• It’s time to retire the term “user”
• Java vs. Scala: Comparative Analysis for Backend Development in Fintech
• Google Announces Agent Builder, Expanded Gemini 1.5, Open-Source Additions
• Chronon, Airbnb’s ML Feature Platform, Is Now Open Source
• US Air Force confirms first successful AI dogfight
• DDoS threat report for 2024 Q1
• The AI Device Revolution Isn’t Going to Kill the Smartphone
• Technology Radar vol.30
• Stanford AI Index Report 2024
  • AI beats humans on some tasks, but not on all.
  • Industry continues to dominate frontier AI research.
  • Frontier models get way more expensive…For example, OpenAI’s GPT-4 used an estimated $78 million worth of compute to train, while Google’s Gemini Ultra cost $191 million for compute.
  • The United States leads China, the EU, and the U.K. as the leading source of top AI models.
  • Robust and standardized evaluations for LLM responsibility are seriously lacking.
  • Generative AI investment skyrockets.
  • The data is in: AI makes workers more productive and leads to higher quality work.
  • Scientific progress accelerates even further, thanks to AI.
  • The number of AI regulations in the United States sharply increases.
  • People across the globe are more cognizant of AI’s potential impact—and more nervous.

• Google Cloud Next ‘24 – a recap
• Development velocity
• Elon Musk’s xAI previews Grok-1.5V, its first multimodal model
• Naval Ravikant’s Airchat is a social app built around talk
• AI Models : “...That’s the most common model for it. And so, today’s models cost of order $100 million to train, plus or minus factor two or three…The models that are in training now and that will come out at various times later this year or early next year are closer in cost to $1 billion. So that’s already happening. And then I think in 2025 and 2026, we’ll get more towards $5 or $10 billion…“
• Big banks are in a race to get AI right…AI Innovation in Banking
• Humane AI Pin review: not even close
• Global internal payment fraud system at Delivery Hero
• Space: The $1.8 Trillion Opportunity for Global Economic Growth
• Fine porcelain maker Maruwa becomes the hottest bet for cooling AI data centres

• Elon Musk announces Tesla will unveil a ‘robotaxi’ on August 8
• Apple changes App Store rules to allow retro game emulators globally
• Varda Space’s orbital drug factory success fuels $90M in new funding
• Sundar Pichai on the challenge of innovating in a huge company
• Engineers Pinpoint Cause of Voyager 1 Issue, Are Working on Solution
• Roku’s New HDMI Tech Could Show Ads When You Pause Your Game
• Denis Villeneuve is doing Dune 3
• Mars may not have had liquid water long enough for life to form
• Google considers charging for AI-powered search in big change to business model
• AI Ascent 2024 – Video Playlist

• Educators divided over AI’s impact
• CodableRPC: An iOS UI Testing Power Tool
• Xiaomi releases electric car $4K cheaper than Tesla’s Model 3
• The Field Guide to Non-Engagement Signals
• Major companies put U.S. cyber defenses to the test in simulated attack
• Axios Review: Using Apple’s Vision Pro in real life
• NYC will test AI gun detectors on the subway
• Model Excellence Scores: A Framework for Enhancing the Quality of Machine Learning Systems at Scale
• Scaling AI/ML Infrastructure at Uber
• 30 books that Elon Musk, Jeff Bezos, and Bill Gates recommend reading
• 20 years of Gmail
• Being a Backend Developer Today Feels Harder Than 20 Years Ago
• How We’ll Reach a 1 Trillion Transistor GPU
• Amazon concludes $4 billion investment in Anthropic
• How Apple plans to update new iPhones without opening them
• Distributed Caching: Enhancing Performance in Modern Applications
• Monitoring Generative AI Applications in Production
• Amazon.com Inc. plans to spend almost $150 billion in the coming 15 years on data centers, giving the cloud-computing giant the firepower to handle an expected explosion in demand for artificial intelligence applications and other digital services.

Post by @benedictevans

View on Threads

• Book Recommendations from the AWS Enterprise Strategy Team
• Google DeepMind’s new AI assistant helps elite soccer coaches get even better
• Formula 1 chief appalled to find team using Excel to manage 20,000 car parts
• GitHub’s latest AI tool can automatically fix code vulnerabilities
• Google Releases Second Android 15 Developer Preview
• Nvidia’s keynote at GTC held some surprises
• World’s first Neuralink patient enjoying online chess, long Civ 6 sessions
• Cloudflare Open Sources Pingora, a Rust framework for Developing HTTP Proxies
• Fostering an Experimentation Culture in Software Development
• OpenAI Courts Hollywood in Meetings With Film Studios, Directors…pitching its new AI video generation tool, Sora
• Google is bringing satellite messaging to Android 15
• Meet Cherry, an AI shopping assistant that helps you discover products using screenshots or images
• GitHub’s latest AI tool can automatically fix code vulnerabilities
• BBC develops AI plans and talks to Big Tech over archives access

• Why Taiwan Is Building a Satellite Network
• Navigating the Industrial Metaverse: A Blueprint for Future Innovations
• Swarms of low-cost drones powered by AI
• PS5 Pro Specs
  • Rendering 45% faster than PS5
  • 2-3x Ray-tracing (x4 in some cases)
  • 33.5 Teraflops
• Software Company HashiCorp Is Weighing a Potential Sale
• A Famed Danish Chef Is Offering High-End Space Dining
• Real estate agents in China turn to Douyin and Kuaishou to show properties virtually through livestreaming, drawing the attention of out-of-town buyers.
• Google says Chrome’s new real-time URL scanner won’t invade your privacy
• Mercedes is trialing humanoid robots for ‘low skill, repetitive’ tasks
• Azure OpenAI’s “Use Your Data” Feature Now Generally Available
• Devin, the first AI software engineer
• Brief History of Scaling Uber

• How Does Throttling Work?
• The AAARRRP Developer Relations Strategy Framework
• Operating a Large, Distributed System in a Reliable Way
• Amazon goes nuclear, acquires Cumulus Data’s atomic datacenters for $650M
• LinkedIn says premium subscriptions bring in $1.7 billion
• Instagram overtakes TikTok in app downloads in race for new users
• China is in the process of raising more than $27 billion for its largest chip fund to date
• Apple’s BNPL Plan Will Show Up on Your Experian Credit Report. Here’s What That Means
• Anthropic Unveils Claude 3 Models, Highlighting Opus and Its Near-Human Capabilities
• The many uses of mini-organs
• AI Prompt Engineering Is Dead Long live AI prompt engineering
• A new tool called Writable, which uses ChatGPT to help grade student writing assignments

• Smartphone makers are dreaming of a ‘supercycle’ driven by AI. Analysts disagree
• Copilot for OneDrive will fetch your files and summarize them
• Chip giant TSMC opens first Japan factory as it diversifies away from Taiwan
• First Commercial Moon Landing Returns U.S. to Lunar Surface…more commercial era for space exploration
• Uber’s CacheFront: Powering 40M Reads per Second with Significantly Reduced Latency
• Humanoid robot startup Figure AI valued at $2.6 billion as Bezos, OpenAI, Nvidia join funding
• Spotify launches a $9.99/mo standalone audiobooks service for its free users
• Photomath is officially Google’s latest app on the Play Store
• Apple Open Sources Pkl, a Configuration as Code Programming Language
• An AI system that offers emotional support via chat
• GitHub enables push protection by default to stop secrets leak
• “…Just a decade ago, data centers drew 10 megawatts of power, but 100 megawatts is common today…enough to power around 80,000 U.S. households (U.S. DOE 2020).”
• Syrenna’s WaterDrone is the ocean-monitoring ‘underwater weather station’ of the future
• Generative AI: Shaping a New Future for Fraud Prevention
• Search suggestions in Chrome
• Apple abandons its car: Here are other projects the company has killed
• “…Security researchers created an AI worm in a test environment that can automatically spread between generative AI agents—potentially stealing data and sending spam emails along the way…”
• “…Code uploaded to AI developer platform Hugging Face covertly installed backdoors and other types of malware on end-user machines, researchers from security firm JFrog…”
• The desert planet in ‘Dune’ is plausible, according to science
• Will NASA be able to return Mars samples to Earth? New audit raises doubts
• Four reasons Walmart wants to buy smart TV maker Vizio

Apple's auto project was always a moonshot. Makes sense to move GenAI up in priority as that will yield more fruit over next 5-10 years than auto projects.

— Ben Bajarin (@BenBajarin) February 27, 2024

ROI on AI:

AI is handling 2/3 of Klarna’s customer service chats with a 2 minute resolution time vs. 11 minutes for previously and a 25% drop in repeat inquiries.

AI is going to have a massive impact on every industry. pic.twitter.com/TY0BUCUKm0

— Gavin Baker (@GavinSBaker) February 28, 2024