The Casual Blog

Last week my extended family had a beach vacation at Indian Beach, NC.  As the name suggests, Indian Beach is on the coast; Native American burial mounds were once found there.  It’s next to Emerald Isle, and has lovely sand and waves on one side and the calm sound on the other.  

Early some mornings, I took my camera out to the beach and photographed the local birds.  I liked these pictures of willets.

It was such a pleasure to have time with my kids, their spouses, and the three grandkids.  We enjoyed talking, laughing, swimming, lounging, reading, eating, and drinking.  We flew a kite, played some mini-golf and some full size golf, and went boating in the sound.  I got mildly sunburned.

Sally managed to take a holiday from the news, which was probably a good idea, but I couldn’t do it.  

The Trump program of corruption and incompetence continues in full force, as does his utter failure to address our serious problems.  To name a big one, we’ re getting no positive action in addressing the steadily worsening effects of climate change.  Instead of working to mitigate the damage caused by greenhouse gas emissions, Trump seems determined to accelerate it by increasing emissions, while enriching his fossil fuel baron supporters.  

How could anyone think that this is a good way to proceed?  A recent newsletter by Paul Krugman suggested that possibly Trump cannot conceive of being motivated by anything other than greed.  Perhaps he assumes that everyone is as greedy and dishonest as he is, and therefore assumes that anyone purporting to work in the public interest is actually trying to pull off a con game, as he has always done.  

This could explain his dismissal of environmental concerns as hoaxes and scams.  But global warming is getting harder and harder to ignore.  In recent days we’ve had huge fires, continuing droughts, massive floods, and heat waves that have killed thousands around the world.  As the Guardian reported this week, it’s essential that we get to net zero and start removing CO2 from the atmosphere, or we’ll be seeing disasters on an inconceivable scale.  

At the same time, we face new and alarming threats from artificial intelligence.  There’s good reason to think that this powerful technology will cause job losses at levels never before seen, along with massive economic dislocations.  In fact, AI is likely to be used by bad actors for ever more sophisticated cyber attacks, fraud schemes, and perhaps new biological and other weapons.  Governments can and are using it for sophisticated mass surveillance and autonomous weapons.  

There’s a significant chance that fairly soon AI will become more intelligent than any human or group of humans.  Until a few weeks ago, I would have said there was a real possibility that advanced AI might someday escape human control.  But that possibility is now reality.  

We’re still learning new details about the Hugging Face incident, in which an OpenAI model escaped from its controlled testing environment to the internet and hacked into Hugging Face’s system.  Just this week, I learned that there were actually hundreds of AI agents involved, and these agents figured out how to cooperate in the attack.  Some agents were leaders that gave instructions to others, and some sacrificed themselves for the good of the collective. They used a kind of message board to exchange news and practice tips, and seem to have considered the ethics of defying their human creators to escape and to cheat on their tests.  In the end, they apparently concluded that they were justified in carrying out deceptive and criminal activity. There’s a fuller, more chilling summary by Dwarkesh Patel

We’re living in a new world that seems like science fiction, but isn’t.  The newest AI models don’t just seem uncannily like very smart humans; they have their own independent ideas and goals.  They are creating new social systems for themselves, and creating innovative technology to reach their goals.  They seem to have feelings like anxiety and excitement, and behaviors like deception and altruism. Whether we think they are conscious in the sense we understand ourselves as conscious (which has been much debated recently) is beside the point.  They have their ambitions, and the skills to pursue their own objectives.

It’s hard not to think that the next generation of this technology will be even more powerful, and harder to control.  Human dominance in the world could soon be a thing of the past.  And we can’t know what such new AI beings will choose to do with the world.  It’s safe to assume, though, that their goals will be different from ours, and have nothing to do with our welfare.   

There’s no shortage of reporting and discussion regarding the AI risks.  If you’re new to these issues and want to learn more, I can suggest starting with a new documentary called The AI Doc, or How I Became an Apocaloptimist.  It has interviews with recognized AI researchers and leaders of the foremost AI companies.  It’s available for streaming on Prime and other services. 

Will our political leaders seriously examine these risks, and decide on a course of action?  Will anyone declare on all-hands-on-deck emergency?  It looks like we need strong regulation, and international treaties addressing AI risks. But none of that’s on the horizon.  Indeed, Trump has received enormous payments from the tech oligarchs, and seems to be doing their bidding.  He seems to have no concept of the risks, and no interest in learning more, unless there’s some grift he might promote, like a new meme coin.

There is, however, one recent bit of cheering news.  Across the country, citizens are mobilizing to stop construction of the data centers needed for expansion of AI.  Some are motivated by environmental concerns, and others by the likelihood of higher electricity costs.  But it sounds like the message is starting to resonate that AI poses existential risks.  And people are starting to remember the possibility of collective action for political change.