- Chaos Theory
- Posts
- 🥟 Chao-Down #115 OpenAI plans a "personal assistant for work", IBM bringing generative AI commentary to Wimbledon, Latest research in building large multimodal models
🥟 Chao-Down #115 OpenAI plans a "personal assistant for work", IBM bringing generative AI commentary to Wimbledon, Latest research in building large multimodal models
Plus, a new study finds that jobs held by women are more exposed to AI risk than men.
Will AI disproportionately affect the jobs of women compared to men? A new report from seems to suggest so.
The Kenan Institute of Private Enterprise recently completed a study and found that nearly 80% of women in the US workforce, compared to 60% of men are in jobs that are more likely to be disrupted to automation through generative AI. Specifically, they defined “highly exposed” occupations as jobs where over 25% of tasks could be automated.
These jobs include office and administrative support, healthcare, education, and social services, with white-collar work more likely to be in danger compared to blue-collar jobs.
It’s an early finding but is an important data point we all should consider as we’re building a future that will be increasingly driven by AI.
-Alex, your resident Chaos Coordinator.
What happened in AI? 📰
Open-source licenses need to evolve to deal with AI (The Register)
OpenAI Plans ChatGPT ‘Personal Assistant for Work,’ Setting Up Microsoft Rivalry (The Information)
The people paid to train AI are outsourcing their work… to AI (MIT Technology Review)
AI Threatens Many Jobs but More Than a Quarter of Workers Still Plan to Quit (Bloomberg)
Robots learn to perform chores by watching YouTube (TechCrunch)
IBM is introducing generative AI commentary to Wimbledon (qz.com)
Always be Learnin’ 📕 📖
A Senior Engineer’s Guide to FAANG Interviews (interviewing.io)
MIT6.S192: Deep Creativity (ali-design.github.io)
When Do Startups Scale? Large-scale Evidence from Job Postings (SSRN)
Projects to Keep an Eye On 🛠
turboderp/exllama: A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights. (Github)
va1shn9v/PromptIR: PromptIR: Prompting for All-in-One Blind Image Restoration (Github)
Giskard-AI/giskard: The testing framework dedicated to ML models, from tabular to LLMs 🛡️🧑🔬 (Github)
The Latest in AI Research 💡
Is ChatGPT a Good Recommender? A Preliminary Study (arxiv)
Large Multimodal Models - Towards Building and Surpassing Multimodal GPT-4 (slides)
The World Outside of AI 🌎
The Particular Misery of College-Admissions TikTok (The New Yorker)
Apple is plotting the death of the password (qz.com)
Charitable giving in 2022 drops for only the fourth time in 40 years: Giving USA report (AP News)
Trauma, gaslighting, boundaries: The limits of therapy-speak (Vox)
Magic Mushrooms. LSD. Ketamine. The Drugs That Power Silicon Valley. (WSJ)