Bruce,
From watching the webcast live, it looked like the “Huh” response was triggered when a Judge sent more than 1 input before Rose responded.

I propose the following hypothesis:

If, from a contest opening round, a 13th place contest entry joins a 1st place contest entry,
then on average (1+13÷2=7), a 7th place contest entry from the opening round is entered
to the final round, as one of the top 4 entries.

Not very logical, since both the 5th and 6th place contestants would be passed over.

The hypothesis may be tilting at windmills.

How was the press coverage this year? Did anybody get an on-camera interview?

Merlin - Sep 18, 2017:

Not very logical, since both the 5th and 6th place contestants would be passed over.

The hypothesis may be tilting at windmills.

I think that was the point. Why not just pick 4 arbitrary entries for the final. If one of those just happens to be Tom’s entry, so be it.

∞Pla•Net - Sep 17, 2017:

1. Contestant A makes one entry.
2. Contestant M makes one entry.
3. Contestant A entry joins Contestant M entry.

Does the CONTEST RULE disqualify Contestant A entry and Contestant M entry?

In this scenario there are still only two contestants and two entries: 2/2 = 1 entry per contestant. We get it, you’re bummed that your interface broke when you unnecessarily reconfigured it to work with both protocols. It’s a moot point to pine over because your bot wouldn’t have made it to the finals either way. This isn’t a contest of who has the better interface, it’s about which bot has the most human responses. Richard did not “join” Steve to edit Mitsuku’s responses, Dave did not join Bruce to edit Rose’s responses, and Denis Robert did not join all of us to edit all our bots’ responses. Most of us are guilty of helping one another out in developing and debugging the interface.

In my hypothesis, the 1st place contest entry acted as a contest judge,
by passing over the places between 13th place and 1st place, to enter
two entries from the qualifying round into the final round. In theory,
the contest rule limit of 4 top entries was broken. Theoretically,
all entries may decide to act as a judge and simply split the prize.

Don,

So you want to compete in the final round against a contestant you already defeated in the qualifying round?  That suggests you did not realize that 5th place contestants enter the final round under my hypothesis. In theory, the 5th place contestants technically never failed. This theory is contingent upon the disqualification of the 1st place and 13th place contestants, or at least letting their average qualifying round performance (scores, place numbers) determine entry to the final round.

Bruce,
From watching the webcast live, it looked like the “Huh” response was triggered when a Judge sent more than 1 input before Rose responded.

Thank you Merlin. That’s the only clue I will have to what happened. To add insult to injury they failed to pull Rose’s log file off the machine before returning it (they did know they were supposed to but became confused by erasing LPP directory data (needed to be done between rounds because the bridge fails to erase subfolders as it reads them).

Robby Garner - Sep 18, 2017:

How was the press coverage this year? Did anybody get an on-camera interview?

http://www.huffingtonpost.com/entry/the-turing-test-is-tired-its-time-for-ai-to-move_us_59bfae52e4b02c642e4a187a

Prize Awarding

from the article - Sep 18, 2017:

...His three victories in the Loebner prize notwithstanding, Worswick is ready for a new challenge. “Only the arrogance of the human race believes nothing can be more intelligent than a person — and I’m sure back in Turing’s day this was the case, but it’s now time to move on.”

Care to elaborate Steve?

Perhaps Steve is working on a Frankenstenian robot in his basement, ready to usher in the robot revolution

Rose’s first “huh?” does immediately follow after another output in every round except 4. It also doesn’t appear to be her normal, capitalised “Huh?” response. I don’t suppose there’s an uncapitalised “huh?” built into the interface somewhere trying to pass for an error message? It’s not in the bridge, at least.

For the sake of conversation, let’s discuss Information Technology from thousands of miles away…

Bruce said, “To add insult to injury they failed to pull Rose’s log file off the machine”

Then just ask them for the Internet Information Services log files.  Windows Start menu -> Administrative Tools -> IIS Manager.

Bruce said, “From watching the webcast live, it looked like the “Huh” response was triggered when a Judge sent more than 1 input before Rose responded.”

A closer look at message-mode reveals it has an input buffer, which may be edited after being sent, and before being received… 

Reference: https://www.gnu.org/software/emacs/manual/html_mono/message.html
“Message will then pop up a new message mode buffer with appropriate headers filled out, and the user can edit the message before sending it.”

In other words, in message-mode the message may be postponed and the message buffer can be opened…
“In order to go back to a postponed message, just open the \*message\* buffer.”
https://www.emacswiki.org/emacs/MessageMode

Character-mode, on the other hand, recorded micro-second the timing of each keystroke.  Perhaps for Turing test competitions, message-mode may be more of a downgrade, than an upgrade?

http://www.huffingtonpost.com/entry/the-turing-test-is-tired-its-time-for-ai-to-move_us_59bfae52e4b02c642e4a187a

Prize Awarding

 

Hey, I took that photo!  They must have got it from Steve.

Professor Wallace,

Hey, the Huffington Post gave Pandorabots a link…

Mitsuku was developed with AIML language and is hosted by Pandorabots.

“Then just ask them for the Internet Information Services log files.  Windows Start menu -> Administrative Tools -> IIS Manager.”

They rented all the machines and returned them.