2002
Selekoa’s 24 records dated 2002, in date order, grouped by topic.
Selective: what the records hold, not a history of the year.
How this page is made
- Each record keeps the kind of date its source gives: an event, a publication, an announcement, an effective date, a revision, a signature, a snapshot, or the start or end of an interval. A snapshot is one dated report of a state.
- A record dated only to a month sits under that month, before the records dated to a day. A record dated only to the year sits under the year, before the months. No day or month is added.
- An announcement records what a party stated. It does not show that anything was done, built or paid for.
- Dates a party states for a year are listed apart, after the records: targets, deadlines and terms, and dates stated in passing, such as a reported start of operation. So is a record’s own date that falls after its sources were last checked, such as a law’s date of effect. A stated date given only in words, with no year, is not listed.
- Records with an unknown date are not listed. A record with an interval appears in the year it starts and in the year it ends, or once, with both dates, when both fall in one year.
- Each provision of a text is a record of its own. Provisions sharing a text and a date are shown here as one entry, with their number, and listed on the text’s page; a contested or unresolved one is listed here on its own. The counts on this page include each provision record.
- The page is generated from the records, with no selection or framing by an editor. Coverage is uneven across years and topics: a year or topic with few records may be one the records barely cover.
Ideas & lineages 24 records
Dated to 2002 only
· publication · dated to the year
“The AI-Box Experiment” (2002) credits Eliezer S. Yudkowsky as its author. Evidence ↗· publication · dated to the year
“The AI-Box Experiment” (2002) contests Y06. Explains the rule that the AI party wins only if the gatekeeper voluntarily lets it out, and names the claim the experiment was meant to test. The page describes two 2002 role-play tests in which Yudkowsky played the AI. Evidence ↗
March 2002
· publication · dated to the day
The mailing-list message “The "AI Box" experiment” (2002-03-08) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
Yudkowsky’s SL4 message of 2002-03-08 proposes a role-play in which the author plays a mind in a box and tries to persuade the other person to let it out. Evidence ↗· publication · dated to the day
The mailing-list message “Re: Novamente goal system” (2002-03-09) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: Novamente goal system” (2002-03-09) contests Y21. Replying to Ben Goertzel on the Novamente goal system, he lists as an anthropomorphic intuition the expectation that Novamente will be 'socialized' by interacting with humans, and argues that human socialization depends on evolved adaptations an AI would lack. Evidence ↗· publication · dated to the day
The mailing-list message “Webmind Inc. as social process” (2002-03-10) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Webmind Inc. as social process” (2002-03-10) bears on Y24. Writing to Ben Goertzel about convincing him of Friendly AI, he says SIAI may simply have to finish its own seed AI first, but that he would not mind another project winning if it understood Friendliness. Evidence ↗
May 2002
· publication · dated to the day
The mailing-list message “Re: supergoal stability” (2002-05-03) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: supergoal stability” (2002-05-03) contests Y01. Replying to Wei Dai in a thread on supergoal stability, he says a non-Friendly mind will keep whatever goals it had when it first became able to protect its goal system from modification, and that only the CFAI semantics would keep goals as the programmers understood them. Evidence ↗· publication · dated to the day
The SL4 message “Re: supergoal stability” (2002-05-03) bears on Y24. Replying to Wei Dai on goal-system stability, he refuses to design Friendliness that would serve a programmer's propaganda and says the only thing to do about a project that tries this is to try to beat it to the punch. Evidence ↗
June 2002
· publication · dated to the day
The mailing-list message “Re: Threats to the Singularity.” (2002-06-23) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: Seed AI (was: How hard a Singularity?)” of 2002-06-23, archive message 4219, is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: Threats to the Singularity.” (2002-06-23) contests Y18. Replying to Ben Goertzel's low-confidence guess that superintelligent Powers might neither help nor attack humans, he says deliberate extermination is unlikely but indifference would probably amount to extermination. Evidence ↗· publication · dated to the day
The SL4 message “Re: Threats to the Singularity.” (2002-06-23) contests Y19. Replying to Ben Goertzel, he separates deliberate extermination, which he thinks unlikely, from indifference, which he says would very probably have the same result. Evidence ↗· publication · dated to the day
The SL4 message “Re: Seed AI (was: How hard a Singularity?)” (2002-06-23) contests Y21. Replying to Ben Goertzel's view that an AI will absorb human thought-patterns through interaction, he argues that human socialization and empathy rest on evolved, hardwired mechanisms and that an AGI will not acquire them merely by exposure to humans. Evidence ↗· publication · dated to the day
The mailing-list message “Re: Military Friendly AI” (2002-06-27) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: Military Friendly AI” (2002-06-27) contests Y16. In a thread on military Friendly AI, replying to Ben Goertzel, he argues that present-day researchers probably hold undiscovered moral errors, so an AI's morality must be able to outgrow its programmers' rather than be fixed. Evidence ↗
July 2002
· publication · dated to the day
The SL4 message “Re: Seed AI (was: How hard a Singularity?)” of 2002-07-02, archive message 4667, is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: Seed AI (was: How hard a Singularity?)” (2002-07-02) bears on Y13. Replying to Samantha Atkins, who questions his earlier warning against waiting until an AI can read humanity's moral literature, he says Friendliness work started at that stage would come too late and must be present from the beginning. Evidence ↗
September 2002
· publication · dated to the day
The mailing-list message “Friendly AI communications issues (was: Consciousness)” (2002-09-10) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Friendly AI communications issues (was: Consciousness)” (2002-09-10) contests Y14. Writing about Friendly AI problems, he declines to rely on the historical pattern that technological problems get solved as they arise and says he wants to know a solution exists in advance. Evidence ↗· publication · dated to the day
The mailing-list message “Re: QUES: CFAI [#2]” (2002-09-19) is sent under the name Eliezer S. Yudkowsky. Evidence ↗· publication · dated to the day
The SL4 message “Re: QUES: CFAI [#2]” (2002-09-19) bears on Y01. Answering a series of questions about Creating Friendly AI, he gives the first of four 'most likely reasons FAI might fail': takeoff is easier than expected, and the mechanisms meant to detect and halt an unexpected ascent fail to notice it. Evidence ↗