We've been measuring this too - the ratio of energy consumption for inference vs train is staggering. We need efficient AI. #efficientAI #projectscience

Remember folks, the cloud is just someone else's computer!

For LinkedIn posts I've *measured* the half life at around 12 hours. It appears to be designed to maximise 'engagement', but really just drives churn and, increasingly, AI-slop.

👍 We need efficient AI - at #projectscience we use a range of algorithms and favour those that are computationally efficient. Less CPU/GPU = fewer $ = faster + more explainable.

Yep - massive disruption on the way. #projectscience #predictforjira

Have you tried ollama? ollama.com

AI is disrupting.

Yes, we have experience. In fact, we've developed apps to improve Jira's usability - in particular Predict for Jira, that uses machine learning / AI to automatically estimate task durations based on past *actual* performance. Have a look at marketplace.atlassian.com/apps/1233542...

I like it - I've long said that Data is a Team Sport - the most powerful behavioural shifts are when the data serves the team, not reporting. #projectscience

I'm interested in data sets on project delivery - all shapes and sizes, whether small agile projects or large programmes. We use data science to derive insights and make predictions on how long future projects are likely to take.