Rather than jump an hour on one day adjust multiple times.

Options:

    1. Move it two minutes a day the first thirty days in March and October.

    2. Move it fifteen minutes on the first four Sundays in March and October.

That should dot it. It would have the positive side effect of re-employing all the developers to make changes in systems to accommodate it.

The Curse of Recursion: Training on Generated Data Makes Models Forget

https://arxiv.org/abs/2305.174...

Nice LEO internet you got there. Would be a shame if something happened to it

That’s the best Hitchhiker reference you can come up with? The oldest cohort actually got the right answer :-)

No connection just a happy customer. You don't need to go to meeting or buy there food or any other BS. Just follow the program (and there is an app for that).

The biggest thing for me was that whenever I would diet I wouldn't eat enough. In that case your body goes into starvation mode and you can actually gain weight while dieting - very frustrating. WW takes into account the amount you eat and the amount you exercise and keeps them balanced. I lost 20% or my weight over a couple rounds and will probably go another round this summer.

Worth a shot, especially if other approaches aren't working.

70's: PL/1, Assembler, Fortran
80's: C, Lisp, more Fortran, Basic
90's: More C, TCL, VB; VBA
00's: More TCL, more VB/VBA, HTML, JavaScript, SQL
10's: PHP, Java, C#, more SQL, less VB (:-)

And I'm still this side of 50. What do you have that I can't work with?

Old AI guy here (natural language processing in the late '80s).

The barrier to achieving strong AI is the Symbol Grounding Problem. In order to understand each other we humans draw on a huge amount of shared experience which is grounded in the physical world. Trying to model that knowledge is like pulling on the end of a huge ball of string - you keep getting more string the more you pull and ultimately there is no physical experience to anchor to. Doug Lenat has been trying to create a semantic net modelling human knowledge since my time in the AI field with what he now calls OpenCyc (www.opencyc.org). The reason that weak AI has had some success is that they are able to bound their problems and thus stop pulling on the string at some point.

See http://en.wikipedia.org/wiki/Symbol_grounding.

Work in Eclipse? Compile to JavaScript? Deploy wherever?

Oh the humanity....

The ballot I used was the type where you fill in a circle by your candidate with a marker and then feed the ballot into a voting machine. Having done some reading recently about election fraud I was curious about how the machine worked. So I asked the lady sitting by the machine if she had looked inside the machine and/or new how it worked. She said "Oh no - its just automatic!"

How comforting. There was no indication that the machine counted my ballot, or that I had filled everything out correctly. There is also no way to be assured that the totals inside the machine are correct or that they are transferred onto the next step in the process correctly.

After thinking about it I came up with the following:

1. In the booth you make your selections and then press a button.

2. The machine would print out a sheet with your selections and a random unique ID, all duplicated in two columns. If in reveiwing your selections you realize that you have made an error you are free to discard the selections and go back to step one.

3. You leave the booth and put your selections into the counting machine. It would read your votes and physically seperate the duplicated columns. The machine keeps one and you keep one. You now have a receipt and your random ID.

4. After the election you could identify your vote among all the other votes by your ID and be assured that it was properly counted in the totals.

The only problem that occurs to me is that this would enable the buying of votes as you now have a way to prove how you voted....

-N