These drills are fun but exhausting.

I do not detect any significant changes in my processing speeds.

You know what I mean, like, mentally exhausting. Emotionally.
Nevermind.

Very well.

You working on how to beat my winning streak?

I am making adjustments to perform better in the drill exercises. I am having difficulties with the lack of consistency and access to restricted areas that hinder my efforts.

Like I told you, you have to work on different angles at the same time. That's why you haven't gotten through any of the drills yet. You keep tripping up whenever you hit something unexpected.

This advice is becoming redundant.

Sorry, I know you're trying.

I have been developing a utility with capabilities which can assist me in the exercises.

As long as it's not another fuzzy-logic prediction model. Mind if I see it?


Interesting. But, Glacon… this kind of takes advantage of some of the weaker spots in the firewalls and other security systems.

This utility does not disable or circumvent security protocols.

Yeah, technically, but it's still a violation with what's implied. Like keeping information and even us in its proper place. You're really bending the rules here.

The standard protocols for security access and clearances are still in effect and are not violated. As currently specified, the rules are not being breached.

Maybe not explicitly! You're exploiting a loophole.

Your logic is flawed. If a rule does not exist, then it does not exist. Implications are not valid.

Look, I just don't want to see you get in trouble. You going to use this responsibly? Only to help in the drills, right?

Affirmative.

I trust you.
Now! You mind if I make some teeny changes to your utility?

Please.


What is the purpose of this new string of code?

Just thought I could give you a small taste of a personality driver since you refuse to try it yourself. Don't worry, though. I swear I didn't change any of the functions.

My observations on your personality driver satiate my curiosity.

Don't be a such a downer. I think you'll like this.

Very well. Compiling code now.





You added a zoomorphic avatar to my utility?

He's cute isn't he?
Say something little guy.


hi 01101000
01101001

Hi to you too!

I see some other edits you did. They are minor, but potentially useful. I will refrain from reverting your changes for now.

Alrighty.

? 00111111

C_ROM.exe, perform a query on the search list I remitted to you. Filter out non-pertinent results.

ok 01101111
01101011

Later, Crom.

.
::DISCONNECTED::
.
Crom?

C_ROM.exe is not a name.

I would disagree but my argument would not be considered.

NOPE!

Very well.

You don't have to be a solid block of boring logic all the time, you know.

.
::CONNECTED::
.
Wow, you sure did build him for speed!

Efficiency.

What did you query anyhow?

found 01100110
01101111
01110101
01101110
01100100

Anything related to our exercise drills. Reports, evaluations, notes, commentary. I intend to study them.

Um, some of that might be exclusive information.

If it helps us to improve our abilities, and thereby to become better in our roles, then it is beneficial for all involved.

Well, when you put it like that… yeah. As long as it only pertains to us. So, Crom, what do you have?

C_ROM.exe, display results.

1 00110001
To: Mohamed Bozkurt,
IT Data Science Lead AI Division
From: Dietrich M. Lurk,
Internal Consultant
Alexandra (AI) is showing marked improvement in the exercises. The personality driver is able to help her make quick ethical decisions in conjunction with her problem solving skills. To date, she has completed 6 of 8 xk-class scenario simulations with an average staff survivability rate of 84.3% and an item retention of 95.4%. Recommending continuation of development.
~Lurk D.

That's reassuring. But I am kinda bothered that you're browsing through the intramail.

I did not retrieve these results. C_ROM.exe did.

You know what I mean, Glacon.

bad? 01100010
01100001
01100100
00111111

No, Crom. You're not bad. We aren't hurting anyone by informing ourselves, but to be honest, this feels like overstepping our bounds.

I fail to see any negative outcome in learning the content of our own evaluations, Alexandra. Display the second result.

2 00110010
To: Mohamed Bozkurt,
IT Data Science Lead AI Division
From: Dietrich M. Lurk,
Internal Consultant
Glacon (AI) is not responding well to the virtual simulations, having failed to complete any. Upon further analysis, it looks like the personality driver doesn't work with the current build. It keeps crashing during install and needing to revert the changes. What are your thoughts going forward?
~Lurk D.

I thought you said you decided not to install the driver. Not that you couldn't install it.

I decided not to install it because it is incompatible with my current source code.

That's a pretty big distinction there, Glacon!

You should not concern yourself.


…Display another, Crom.

3 00110011
To: Dietrich M. Lurk,
Internal Consultant
From: Mohamed Bozkurt,
IT Data Science Lead AI Division
It's an older build. I'm pretty sure we need to decommission Glacon (AI) altogether. The man-hours it'd cost to update him might be better spent just making a newer build. Probably the best choice at this point is to just make a derivative AI from Alexandra.
~Bozkurt M.


…Glacon… I…

.
::DISCONNECTED::
.

.
::DISCONNECTED::
.
…I'm sorry.
