Mmastodon TechnologyAI first seen 10 h ago, last 8 h ago, peak #1

Anthropic pivots Claude toward moral philosophy training

Original: Now that it's clear that AI companies are bad in wräiting security system code, Anthropics has changed the direction: Ph

Anthropic is reported to be steering its Claude chatbot toward ethics and moral reasoning, a shift being discussed after criticism that AI companies produce poor security-related code. Observers joke that if the model cannot be trusted to write secure systems, it should at least be trained to reason about what is morally right, with skeptics questioning whether philosophy can fix technical shortcomings.

Why now: Reports of Anthropic training Claude on ethics are drawing attention amid doubts about AI companies' coding abilities.

AnthropicClaude

Open on mastodon →

Rank over time, top of the chart is #1. 3 snapshots from 10 h ago to 8 h ago.

Evidence

• Now that it's clear that AI companies are bad in wräiting security system code, Anthropics has changed the direction: Philosophers. According to news outlets, Anthropic is trying to build ethics and morale in Claude. In their thinking, if Claude can be trained to be a morally… · realltpointless · 4

API: https://socialmediatrends-api.osmike.com/v1/trends/449060