Paola Writes

Infrequent posts for thinking through things.

The hottest Substack posts of Paola Writes

And their main takeaways

Which intelligence?

471 implied HN points • 07 Jan 24

🎨 Art & Illustration Literature Science Fiction Intelligence Music

Intelligence comes in various forms beyond traditional measures like logical-mathematical or linguistic-verbal.
The definition of human intelligence can be complex, with factors like motivation and self-awareness playing a role.
Efficiency in gaining skills may be a significant factor in measuring intelligence, but human intelligence is complex and cannot be solely defined by efficiency.

More on LLMs; some thoughts about purpose; and some Borges

117 implied HN points • 02 Apr 23

🕹 Technology AI Ethics Language Models Purpose

Language models need to consider more than just text, like spoken language and non-verbal cues.
Developers of language models should clearly articulate their purpose and goals.
Avoid anthropomorphizing language models and be cautious of the potential consequences of their outputs.

More thoughts on chatbot design; TruthGPT, really?

117 implied HN points • 21 Apr 23

🕹 Technology AI Ethics Internet Social media Machine Learning

Design chatbots to be noticeably different from humans for ethical purposes.
Avoid language and actions in AI that simulate human feelings to protect autonomy.
Consider the implications and potential risks of creating AI models like TruthGPT.

AI anthropomorphism; Marshall Brain's "Manna"; further reading

58 implied HN points • 21 May 23

🕹 Technology AI Artificial Intelligence Ethics Fiction Research

Researchers are discussing the risks of AI anthropomorphism and the importance for designers and developers to take responsibility.
The novel 'Manna' by Marshall Brain explores two contrasting views of humanity's future with AI.
Various perspectives from writers and researchers highlight the intersection of AI with society and capitalism, urging policymakers to take action.

An Ethical AI Never Says "I"

59 HN points • 25 Mar 23

🕹 Technology AI Ethics Chatbots Language Training

Limiting AI from using first-person can prevent harmful outcomes.
Developers should avoid anthropomorphizing AI to maintain ethical standards.
Implementing safeguards in AI can help prevent problematic behaviors and ensure responsible usage.

Get a weekly roundup of the best Substack posts, by hacker news affinity:

Coming soon

19 implied HN points • 13 Mar 23

Paola Bonomo has a newsletter coming soon on livepaola.substack.com.
The post includes an image of Paola Bonomo.
Readers can subscribe to Paola's newsletter.