You are right.
AI systems are much easier to align because we get to *design* their intrinsic objectives.
We don't get to do this with humans. We only get to modify objectives slightly through education.
SYSTEMS
-
Designing AI Objectives: Alignment Challenge vs Human Nature
By
–
-
Healthcare.gov State Machine Error: Citizenship Status Misclassification
By
–
An example of "I certainly understand why you have architected system to work this way, but NO PERSON WANTS IT TO WORK THIS WAY" from the healthcare dot gov state machine, which wrongly came to believe Ruriko was a citizen then got confused. I sent in green card. Note bold bit.
-
Distributed Systems in 9p Bootcamp Initiative
By
–
yes, but it's still kicking around. i'm using a distributed system as part of the http://
9p.sdf.org bootcamp -
Algorithm Reversibility: Control in Autonomous Decision-Making Systems
By
–
Is the decision your algorithm makes reversible? e.g. light switch vs potentially lethal autonomous car. This can impact how much control a user feels they have. Insights from this great talk on causality, from a philosopher of neuroscience.
-
T-Systems Launches Digital Twin Offering with NVIDIA Omniverse
By
–
T-Systems launches a digital twin offering with @nvidia Omniverse https://actuia.com/actualite/t-systems-lance-une-offre-de-jumeau-numerique-avec-nvidia-omniverse/ … #AI #artificialintelligence @tsystemscom @TSystemsFrance
-
Manufacturing Data Superiority: Three-Phase Smart Optimization Strategy
By
–
Manufacturing and Data Superiority in 3 phases: – Extracting insights from operational data for decisions. – Using historical data to forecast trends. – Implementing smart systems that self-adjust, optimizing operations based on continuous data feedback. Microblog @antgrasso
-
Google Napa: Progressive Partitioning Optimizes Reporting Queries
By
–
Google Ads infrastructure runs on Napa, an internal data warehouse that serves billions of reporting queries each day. Today, learn how a progressive partitioning algorithm efficiently optimizes database queries to meet strict latency targets. Read more ↓
-
Air-Guardian: MIT AI System Corrects Pilot Errors
By
–
Air-Guardian' is an @MIT_CSAIL system for aviation that uses AI to correct pilot errors. Tracking attention through eye-tracking and saliency maps anticipates risks compared to traditional autopilots, and promises improved safety.
— Antonio Grasso (@antgrasso) 20 octobre 2023
Link > https://t.co/52LJci9xFk via @antgrasso pic.twitter.com/hCBHVYBTA7Air-Guardian' is an @MIT_CSAIL system for aviation that uses AI to correct pilot errors. Tracking attention through eye-tracking and saliency maps anticipates risks compared to traditional autopilots, and promises improved safety. Link > https://
bit.ly/3Zv5kPj via @antgrasso -
Composable Data Systems: 15 Years and Future Outlook
By
–
Wes McKinney – The Road to Composable Data Systems: Thoughts on the Last 15 Years and the Future https://
bit.ly/3ZgbqCQ
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Automation Reduces Administrative Delays Through State Machine Management
By
–
“Why does that matter?” Because it lets computers do the inglorious-and-unworthy work of administering a state machine rather than having multiple delays introduced by various parties having multiple people involved in reading email, checking bank records, etc.