Bubbles

 (I have a sneaking suspicion I've written about bubbles before.  Apologies if this is repetitive, but I was inspired by an episode of Dessert Masters).

The first economic bubble ever recorded was Tulipmania (1636–1637).  Rare tulip bulbs in the Netherlands became wildly popular, and the market collapsed just as quickly.  Sales in the year after the bubble went down by 99%.

There was another economic bubble in the 1840s when everyone started investing in the railway, which overinflated its value.  Similarly, in 2000 the dot-com bubble over-valued online companies, and the dreams of instant wealth crashed very quickly.

So...is AI a bubble?  

Companies making and selling AI certainly don't think so.  Scaremongers assure us that it will change the world (for the worse) within ten years.  Those very smart people who analyze trends keep shaking their heads with grave concern for people's jobs.  And, worst of all, I know of many people whose employers have happily embraced AI and terminated their employment in areas such as journalism, translation and administration.

I'm also studying AI through an apprenticeship, and I have discovered something.  Without DETAILED instructions, SUSTAINED maintenace, and SKILFUL manipulation of data, AI can't come up with anything better than this.

I hope that employers soon realize that people are more useful than AI.  More trustworthy.  More complex and beautiful and helpful and creative. Even if AI becomes ten times more powerful and accurate than it is now, it can still never match a human.  After all, where did AI get all its data and structure and powerful network?

People.