Summary of recent tweets

H. Sebastian Seung has been tweeting about various topics lately, including AI safety, neuroscience, and advancements in the field. In one tweet, he questions the board's response to a power struggle over AI safety, suggesting that there may be undisclosed matters involved. He expresses skepticism and finds it fishy.

In another tweet, Sebastian Seung shares a collaboration effort called SegCLR that led to surprising capabilities in identifying neuron types from small fragments of wiring. He acknowledges the outstanding effort of his collaborators and highlights their collaboration with the Lichtman lab at Harvard and the Allen Institute.

Seung also discusses the ongoing digital transformation of neuroscience and its potential impact on advancing the field. He mentions software eating neuroscience and suggests that it will take time to unfold but is already underway.

Furthermore, he shares exciting developments in his own research involving fly neural circuits. Through analyzing connectomes without recording neural activity directly, Seung discovers new circuits related to visual perception in flies. He compares these findings to mammalian visual cortex cells and emphasizes their surprising connectivity motifs.

Overall, H. Sebastian Seung's recent tweets cover a range of subjects such as AI safety concerns, collaborations in neuroscience research, advancements in understanding neural circuits using connectomics methods, and the ongoing digital transformation of neuroscience.

Regarding new trends in AI specifically mentioned by H. Sebastian Seung:

1. The digital transformation of neuroscience
2. Advancements in connectomics and analysis of neural circuits
3. The use of machine learning techniques for identifying neuron types from small fragments

An overall sentiment analysis indicates that H.Sebastian Seung's tweets contain mixed sentiments towards AI. While expressing skepticism about certain situations related to AI safety, he also showcases excitement about recent advancements in understanding neural circuits through innovative methods like connectomics and machine learning. Therefore, his sentiment appears to be neither entirely positive nor negative.