This is a pretty good test, how long does it take an AI to write a proper pacman clone, and how much prompting. All of these quick efforts look dismal. So how long would it take? 2 days and a lot of back and forth, something like that? The ghost movement in pacman was quite subtle and would take a long time to get right I reckon.
Also this is hilarious:
Unfortunately, my invitation to chat further over Zoom apparently means I’m after his crypto. “[Expletive] you, scammer,” he says. I ask if it would help if I sent over my Guardian credentials. “How about [Expletive] you?” comes the reply.
Agreed. I feel like the right test would be: "OK, you can use AI to easily generate these pac-man prototypes in an hour or two. But once you get past the 'wow I was able to get all this with so little work', you still have a basically unplayable prototype. Nobody would seriously want to play these or pay for them. How about you make a full-blown pac-man clone with all the nuances of the original? How long does that take? Do you even start converging to the solution at some point or do you keep playing whack-a-mole with bugs and issues?"
Someone builds an access database and it works for a bit but then they leave and no-one else knows how it works so they start exporting the data to excel instead. Access is all about restrictions (the form has these fields; the data is structured like so; input will be validated by these rules) ... Restrictions deliver some value but are brittle - its about as hard to maintain as any other custom built app would be.
as an anecdote, my own mother, who studied french in university and works in education, has built access DBs for places she's worked before. It seems its a very approachable and beginner friendly UI for non-programmers, somehow.
Thats fascinating. Wolf Hall has a lot about Thomas Moore in it ... I should note Wolf Hall is essentially fiction but largely based on things that did happen - I guess you can view it as "lets imagine how the story of Henry VIII would work if much maligned Thomas Cromwell was actually the good guy"
... anyway in Wolf Hall, the character of Thomas Moore as written is largely consistent with what the OP is finding in that old manuscript - someone quite keen on their own cleverness and relatively comfortable with interrogations and burning people at the stake. In Wolf Hall his death is stubborn and needless, and in defiance of the wishes of his wife and daughter. At first I took those parts of Wolf Hall as an exercise in "lets see if its possible to invert the plot of A Man For All Seasons". But then this document "A dialoge concerning heresyes" seems to actually back up the Wolf Hall picture of Moore.
As an aside, the "inversion of A Man For All Seasons" aspect is brilliant. The scene of More and Cromwell together in the Tower of London has this incredible exchange where Cromwell predicts that their dispute will be replayed throughout time and he fears he is being already typecast as the villain. I don't have the book in front of me, so I'm likely misremembering it. But the way that it tips its hat to the play was really moving to me.
I'm obviously preaching to the choir, but damn, Hilary Mantel was brilliant.
There's actually some quite suprising stuff in that article, e.g. some cultures have earlier childhood memories than others, and there's some interesting suggestions as to why that is. Also considerations about how early events can apparently be stored but not consciously remembered.
"""NEW YORK, NY – A recent study showed that 70% of people actually never read more than the headline of a science article before commenting and sharing. Most simply see a headline they like and click share and make a comment. A recent study showed that 70% of people actually never read more than the headline of a science article before commenting and sharing. Most simply see a headline they like and click share and make a comment.
The most frustrating iteration of this is when someone uses an article to make a point in an argument...and they didn't read it! A clickbait headline seems to support their claim, but the substance refutes it.
Not sure if your pun was intended, but 'wait' probably works so well because of the models being trained on text structured like your comment, where "wait" is followed by a deeper understanding.
Also this is hilarious:
Unfortunately, my invitation to chat further over Zoom apparently means I’m after his crypto. “[Expletive] you, scammer,” he says. I ask if it would help if I sent over my Guardian credentials. “How about [Expletive] you?” comes the reply.