The low level of understanding as to the functioning of LLMs and ML continues to allow 'AI' to be publicly imagined as an entity, a being - an idea bolstered and proliferated through countless works of popular fiction and shoddy journalism. This is a vulnerable state for the public imagination, ripe for abuse, especially when coupled with the still prevalent myth of computational neutrality. As such I believe the unchecked spread of 'AI' needs to be fought as much with education as regulation.

The Antarctic Peninsula is rapidly greening, with moss ecosystems showing the fastest extension of their range: https://www.nature.com/articles/s41561-024-01564-5 Hazarding a guess that the next decades will see a tense race to settle the peninsula, and it may not only be those that lay claim (UK, Chile, Argentina). West Antarctica may be next in line, looking at the current shelf melt trends.

So now Netanyahu's government is killing innocents in another sovereign state. When will the international community declare this man fully rogue and hold him and his men accountable for their war crimes and terrorism? How many US and EU weapons are being used to kill Lebanese people?

Nearly 500 killed in Israeli strikes, Lebanon says, as fears of escalation grow – as it happened | Israel | The Guardian

Lebanese health ministry says more than 1,650 people injured; Israel says military moving into next stage of targeting Hezbollah combat structure

We're told we "need" 'AI'/ML just like we were told we needed blockchains, VR, Web3 and the so-called metaverse - these solutions looking for problems. The difference with ML however is that it does solve some big problems very well, the most significant of which (and primary investment driver) is the Late Capitalist project of decoupling skill from People.

Today a dawn trip high up through native beech ('tawhai') to a quiet world of moss set still in cloud. Trees of soft islands drank the air above a teeming carpet of life. #aotearoa