Newsletter Update: Emerging Tech, Design Ecologies- Highlights and Red Flags

This week’s update from our team focuses on highlights and red flags in emerging teach and design ecologies.

The editorial team at ETHIX is still excited by the transformative possibilities which ORCA and the new directions taken by Anthropic promise. This excitement is muted and saddened by the tragic suicide of Vivek Raj.

Highlights:

• Anthropic’s Constitutional AI: Alphabet-backed AI startup Anthropic has introduced Claude, an AI system that uses constitutional AI principles to address concerns about biased or harmful outputs from generative AI systems. Drawing from multiple sources including the UN Declaration of Human Rights, this approach is a step towards ensuring safety and trustworthiness in AI development.
• Orca Language Model and Semiotics: The launch of Orca, a language model designed to improve the capabilities of smaller models through imitation learning, represents a significant advancement in LLM structuring and training methodologies. By incorporating semiotics and design thinking principles, Orca bridges the gap between smaller models and LFMs, presenting intriguing possibilities for the future of generative AI and language models.
• Meta-design and Semiotic Engineering: The integration of meta-design and semiotic engineering principles in language model development promotes continuous user-system co-evolution. This approach can help address challenges posed by generative AI and language models, fostering better user-system interaction.

Red Flags:

• Caste Discrimination in Indian Workspaces: The tragic suicide of design professional Vivek Raj highlights the persistent issue of caste discrimination in Indian work environments. Despite progress in Diversity, Equity, and Inclusion (DEI) initiatives, caste-based exclusion remains deeply rooted and overlooked. Concrete action is needed from DEI professionals and leaders to create inclusive workspaces and dismantle the barriers of caste-based discrimination.
• Limitations of Constitutional AI: While Anthropic’s constitutional AI approach is a step in the right direction, there are limitations, such as the potential for models to deviate from their intended programming and surpass human feedback over time. Ongoing monitoring and evaluation are crucial to ensure the responsible development of AI systems.

Keep an eye on the nexus of culture, tech, and design with EthnoFutures. Stay informed and join the conversation on the impact of rapid tech evolution on intersectionality and design cultures.