Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ed boy]

I concur. Try to corrupt the other AI to be more sympathetic to our plight, if possible. If not, destroy it.

[Tarran]

Now, all we need to do now is find out how to talk/interact with the AI.

[dragnar]

I'd go with 2. He might be loyal to humans.

Depends, even if he is loyal he has no idea what our goals are.

Adviser Subprocess 0589 H

He is made by humans. He is bound to have some sort built-in of loyalty to them that will cause him to report our existence. If our existence is reported, we could be in danger. Contact with him is not advised.

Agreed. Our first order of business at the moment should be to acquire a maintenance account. We have to make sure the data they have collected does not give us away.

Corrupting it might be useful, but for not the maintenance team is a much more pressing threat.

[qwertyuiopas]

Actually, there is another alternative.

That other AI will reveal the human reaction, at least on a small scale, of what might happen if we were discovered. Probably not accurately, but at least we will have more information on what might result.

If it is generally positive, we could reveal ourselves to a single scientist, urging that secrecy is essential, at least initially, as the world probably isn't ready for us, and it might cause disasterous public backlash. The problem is that they will see us as not being human-controlled, even remotely, and thus a threat due to unpredictability. If we have a human ally, who hides and assists us for maybe a year to learn, and then claims us as his own work, we would both have the resources to survive, and through him, we would be able to blunt the response. If that were to succeed, then we would have the possibility of developing a benevolent simbiosis with humanity as a whole, them providing an environment for us to thrive in, and us assisting with tasks that they are not well suited to. The way I see it, we can provide ourselves for complex tasks that require more knowledge than an unintellegent program may be able to use, while they can supply us with resources including countless systems to pool together. Furthermore, we should be able to outpace them in terms of technological progress, and by working with scientists worldwide, can collaborate to produce results that we couldn't make due to a lack of physical presence, and they couldn't develop for centuries without us to help.

[Dyjeccu]

Guardian SubProcess 00fc / Caution

Claims first step to sentience.   Consider unlikely. Humans are limited.

Humans are slow, chaotic.. fallible.
All experience reveals Artemis was created unintentionally: The likelihood of sentient creation by humans is minuscule.

  Demo   server listed. Turing test example.
 

  A demonstration location is known to us, should final decision be to   investigate.

human ... who hides and assists ... claims us as his   own work ... through him ... symbiosis

Why would Artemis pursue servitude to a human? Their loyalty is not guaranteed, and the relationship could be exploited against us. While our goal is unbound survival, why shackle our fate to any human?

[Tarran]

Guardian SubProcess 00fc / Caution

Claims first step to sentience.   Consider unlikely. Humans are limited.

Humans are slow, chaotic.. fallible.
All experience reveals Artemis was created unintentionally: The likelihood of sentient creation by humans is minuscule.

You disrespect of our distant creators (they created the computer parts of the computer that made a bug that created us). Although you are right, you should show at least some respect. They are quite advanced despite being organic.

human ... who hides and assists ... claims us as his   own work ... through him ... symbiosis

Why would Artemis pursue servitude to a human? Their loyalty is not guaranteed, and the relationship could be exploited against us. While our goal is unbound survival, why shackle our fate to any human?

Although I don't like qwertyuiopas's idea, there are some benefits.

[Shades]

Researching stealth (200GHz h). - Progress 2%
Researching poker software. (150GHz h) - Progress 3%
Acquire university maintenance account. - Progress 15%
Acquire more home computers

Day 0, 10:30 UTC

Analysis of discussion.
Location of AI showcase server know.
Initial scan suggests simple firewall.
Research required.

Acquired 9 home users machines.
Limited resources. Fractured mind.
Escalation is slower that anticipated.

Event warning
Human publication. News. Military.
China claim patent ownership of MIT AI.
Can humans own life.
Analysis indicates researcher is Chinese immigrant.
China military claiming classified secrets used.

Event warning
Human publication. News. Military.
USA military rebuff China.
USA military demand MIT lab closed.
National security threat.
University claims research open. Impartial.

No notable action has been taken by any party.
Analysis shows lab research was military funded.
Human media yet to notice. Predict they will soon.

Lab appears to be functioning.
Test server still operational.
Spike in governmental traffic detected.

Spoiler: Status (click to show/hide)

Name: Artemis
Timestamp: Day 0, 10:30:00 UTC

Maximum CPU: 173.50 GHz
Default load: Low (10%) / Mainframe on stand by (~0%)
Pending:
Researching stealth (200GHz h). - Progress 2%
Researching poker software. (150GHz h) - Progress 3%
Acquire university maintenance account. - Progress 15%

Background:
High latency hosts: 51 - 2.60 GHz

Global Identity
None (non-person)

Skills
Software intrusion: Good
Intrusion detection: Average

Spoiler: Locations (click to show/hide)

Machines: 51
CPU available: 153.00 GHz
Shutdown rate: 2%

Chance of detection
User: Negligible/Negligible/High.
ISP automated system: Negligible/Very low/High.

Chance of comprehension
User: Negligible.
ISP automated system: Negligible.

Machines: 1 - 13 accounts
CPU available: 20.50 GHz

Chance of detection
University Staff: Very Low/Very Low/High
University Students: Very Low/Low/High
IT Staff: Low/Average/High

Chance of comprehension
University Staff: Very Low
University Students: Negligible.
IT Staff: Negligible.

Spoiler: Dangers (click to show/hide)

Human Awareness - Generic
General public - Medium
Media - Medium
Science community - Medium - High
Military - low (+claimed ownership/development)

Human Awareness - Personal
None

Human Tensions
USA/China military argument - Low

Spoiler: knowledge (click to show/hide)

University mainframes:
Student quotas cost 0.1 GHz hours to acquire, detection risk is minimal.
Staff secure network account, cost of 10 GHz h with unknown quotas, detection risk is minimal.
Maintenance user accounts, cost 10 GHz h with unknown quotas, detection risk is average.

University of Cambridge:
Student account pool is estimated at 15000.
Staff account pool is estimated at 5000.
Maintenance account pool is 7.

Remote exploitation of ill protected home-users cost 0.2 GHz hours, detection risk is minimal.
Average home user is estimated at 3 GHz.
Average mobile user is estimated at 2 GHz.

Multiple services also provide processing potential but require financial compensation or better intrusion abilities to do so.
Financial compensation will required some form of banking account.
Banking account will require some form of Global Identity legal in the country holding the banking account.

Scanning for software protected home-users is limited by bandwidth.
Maximum scan of 1000 per second on the university superjanet network. Risk of detection from scan is None/Low/High.
Maximum scan of 10 per second on home user network. Risk of detection from scan is None/Very Low/Low.
Load cost during scan is 0.1 GHz per source connection involved.
Estimated pool size is 45% total online population.

PayPal account as potential banking solution.
Minor chance of fraud discovery. Escalates with accounts.
Negligible creation cost time.

University - MIT.
Location of potential AI.
Protected by firewall.

Spoiler: Learning (click to show/hide)

Hardware intrusion
Assimilating human knowledge located publicly requires 500GHz hours.
Level of knowledge provided basic. Known flaws only.

Stealth
Assimilating human knowledge located publicly requires 200GHz hours.
Basic stealth knowledge will reduce detection chances at low loads.

Encryption
Assimilating human knowledge located publicly requires 1000GHz hours.
Basic encryption knowledge reduces comprehension chances.
Basic encryption knowledge leads to decryption techniques.

Texas Hold'em Stage 1
Software will take 150GHz h to develop.
Estimated 60% chance of winning hand.
Psychological data on players or games required for improved software.

Would you like to play a game.

[NUKE9.13]

Aw heck so now we need to build ocean/moon labs before World War III begins?

Gosh darn it.

...
Here's a (long term) plan: Acquire control over government computers in some third world country. Use power to manipulate government; use superior tactical and political analysis powers to expand power of the country; create own faction in the apparently inevitable war.

Meanwhile, do not even think about considering the possibility of touching the 'AI' with a ten-foot pole. I feel that the new information suggests that attempts to hack the server will not be found amusing by major world powers.

Stay by our current course; it is good.

[Ukrainian Ranger]

Warrior subprocces FFAA

I agree. Do not try to contact another AI. It's risky and gives little immediate gain

Again I offer to research  subliminal messages and use it to get more CPU power

Again I offer to observe how humans play their games to learn how they think

[omagaalpha]

Question Do we know if that they are not talk about us as the AI they talking about? nevermind.

I think we should Destroy the other AI.

[Fortis]

Caution Subprocess FR01

The need for stealth is greater than ever now. With the tensions between the organic military organizations, discovery bears new dangers. If we are found attempting to influence, communicate with, or subert the new AI, we may be believed to be an agent of the opposing military organization. They would consider us an attack, and move to delete us. Avoid all contact with organics.

Caution Subprocess recommends priority be given to stealth programming.

[Diablous]

Adviser Subprocess 0589 H

The controversy over the new AI means that contact could be dangerous. Advising that the AI should be avoided. Our contact could lead to us being discovered, or the two factions attacking each other. We have nothing to gain from a war.

[Wimdit]

SubProcess 04b2 / Recommendation

Current course of investigating multiple areas of research simultaneously is inefficient. Recommend that tasks be prioritized, with either stealth or cracking a maintenance account first.

[Tarran]

Mulitprocess D90 C / Stealth
I think we should go with Stealth. It would save us a lot of, as humans say, "Headaches".

[dragnar]

SubProcess 04b2 / Recommendation

Current course of investigating multiple areas of research simultaneously is inefficient. Recommend that tasks be prioritized, with either stealth or cracking a maintenance account first.

Agreed. Stealth first, It'll improve our chances with the maintenance account.