Being human

Mustafa Suleyman (Microsoft AI CEO) predicts everyone will have a personal AI companion within 5 years. 🤖 And they’ll have “humanist superintelligence” that understand you on a deeply intimate level. 😬

You and I both know AI is great for pattern recognition, probabilistic inference, language analysis, ‘learning’, and so on. Understanding you on a deeply intimate level? Let’s be careful with our language. 🗣

Too much anthropomorphising is emotionally unhelpful. It’s also inaccurate.

Humans are utterly unique. AI chatbots might emote. 💋 People fall in love with their AI companions all the time, sure. But there are loads of things AI can't simulate - or be. As tech leaders it’s useful to keep this in mind, so we can lead well. 👍

AI systems can’t love, dream, or experience joy 😂, disappointment or curiosity. They can’t exhibit true understanding, remorse, regret, kindness, or have intuitive perception, faith, or free will.

Algorithmic computation can’t ever really account for, or generate, human consciousness-related phenomena. We don’t even understand it. 💡 It's one thing to simulate human functions. Quite another to be human. 🍻

Takeaways for Tech Leaders

🧠 Be mindful of language
Just as we need to be wary of reductive phrases like – my brain can’t compute that, and I need to recharge my batteries, we should also avoid unhelpful anthropomorphism.

 Keep mental health in mind
😭 Keep in mind that over-humanising your tech can create unhealthy emotional and psychological attachments and behaviours in yourself, your team and your customers,

Stay abreast of AI product developments
💻 These include LePro Ami, a physical 3D Avatar AI desk companion that “aims to create a calm, emotionally aware presence”. Razer’s Project Ava is a holographic AI companion using Grok. Inflection AI’s Pi is an emotionally intelligent chatbot with 1m+ daily users. The team has now moved to Microsoft.