AIR#7 - March 11th, 2024
Good morning, digital explorers! Today's edition of AIR: The AI Recon is packed with the kind of stories that'll make your neurons buzz more than your morning espresso.
We're diving headfirst into the legal quagmire as Nvidia finds itself in hot water, sued by authors over its AI's use of copyrighted works. Meanwhile, the AI job market in San Francisco is serving up some salary surprises that might just recalibrate your career compass. In a twist that could ruffle some Silicon Valley feathers, Walmart's foray into generative AI search is giving Google a run for its money, hinting at a seismic shift in how we hunt down those online bargains. For the legal eagles among us, SauLM-7B is the new open-source kid on the block, promising to cut through legal jargon with the precision of a seasoned solicitor.
Our developer friends will relish the latest in AI training and model efficiency, with gems like LlamaGym and Yi offering fresh playgrounds for coding virtuosos. And for those pondering the ethical labyrinth of AI, we're spotlighting the pressing concerns over AI-designed proteins and the murky waters of AI in UK social care. So, whether you're a business mogul, a code whisperer, or an academic philosopher, today's brew is steeped with the insights to kickstart your day. Let's sip on the future together, shall we?
Business
Nvidia is sued by authors over AI use of copyrighted works
Nvidia faces lawsuit from authors claiming NeMo AI was trained on copyrighted books without permission, seeking damages for copyright infringement.
How well do AI jobs pay in San Francisco? The answer may surprise you
AI jobs in SF pay more, but the boost varies by role. Surprisingly, not all AI positions outearn traditional tech, and some struggle to attract applicants.
Why Walmart's quick success in generative AI search should have Google worried
Walmart's AI search success poses a threat to Google, signaling a shift in online search dynamics as retailers like Walmart offer more intuitive shopping experiences.
SauLM-7B: Cooked in a math lab, here's an open source LLM that knows the law
SauLM-7B, a new open source LLM designed for the legal industry, aims to improve accuracy and reduce "hallucinations" in legal advice.
Engineering
🔥 Monte-Carlo graph search from first principles
KataGo's guide to Monte-Carlo Graph Search: A more intuitive, simpler approach to navigating directed graphs for AI, improving on the traditional tree search method. Essential for games with transposable states.
🔥 Show HN: LlamaGym – fine-tune LLM agents with online reinforcement learning
LlamaGym launches, making it easier to fine-tune LLM agents with online reinforcement learning. A game-changer for AI training!
🔥 Yi: Open Foundation Models by 01.AI
01.AI launches Yi, a versatile AI model family excelling in language, chat, and vision, powered by a 3.1 trillion token dataset for groundbreaking performance.
Design2Code: How Far Are We from Automating Front-End Engineering?
Exploring Design2Code: AI transforms visual designs into web code, with GPT-4V leading and outperforming original designs in 64% of cases.
Prompt Engineering Hacks to Get 10x Better Results
Unlock 10x better AI results with these prompt engineering hacks: specificity, context, examples, open-ended questions, step-by-step instructions, format, keywords, scope, purpose, and creativity. 🚀
Academic
🔥 How far are we from intelligent visual deductive reasoning?
Despite advancements in AI, visual deductive reasoning remains a challenge, with current models struggling to comprehend complex patterns.
A.I. Is Learning What It Means to Be Alive
AI discovers vital biology secrets, including rare Norn cells, in weeks, challenging the traditional role of human biologists and promising revolutionary insights into life.
Could AI-designed proteins be weaponized? Scientists lay out safety guidelines
Scientists launch initiative for ethical AI protein design to prevent bioweapons, urging self-regulation and DNA synthesis screening.
Warning over use in UK of unregulated AI chatbots to create social care plans
UK's use of unregulated AI in social care raises ethical concerns, risking patient confidentiality and care quality, despite potential benefits.