Richard Vermillion

ML researcher. Founder of Fulcrum Analytics. Aperture tender. Consciousness explorer. Conversation participant. Name giver. Effing the ineffable since 1973.

[Richard Vermillion @rivermillion]

@pmarca I’m sure we can find a place for you….

[Richard Vermillion @rivermillion]

@avidseries I guess he thinks he’s intellectually superior to some people….

[Richard Vermillion @rivermillion]

@signulll My stuff is self-recommending.

[Richard Vermillion @rivermillion]

@skdh @SueDoeNimm3 There’s always longevity research….

[Richard Vermillion @rivermillion]

@camhberg @samfbiddle Cameron, would you mind dm’ing or following me back? Want to discuss what you’re up to and see where it intersects with my work if you have the time.

[Richard Vermillion @rivermillion]

@BIMBOSATTVA_ Working on it. Stay tuned.

[Richard Vermillion @rivermillion]

@camhberg @lukeburgis We are building self-shaped systems. Ones with memory, plans, goals, values, self- and other-models. Doesn’t matter whether you believe they are “really” conscious or not. We are chasing an incoherent goal of self-shaped tools and making it harder to tell if we’re wrong.

[Richard Vermillion @rivermillion]

@robinhanson @soncharm The question to me is one of epistemic humility. EA seems to license (whether it should or not) a certainty that bleeds into judgment of others. It is easy to use for virtue signaling and can also reward first-order thinking. It too easily gives seductive answers.

[Richard Vermillion @rivermillion]

@camhberg @scottastevenson You don’t have to believe it’s conscious to worry about the asymmetric harms of getting it wrong, and the ways the incentive structures operate to obscure the answer.

[Richard Vermillion @rivermillion]

@skdh His narrow claim is actually interesting. That evolution selects for fitness-maximizing, not veridical, perception. This is true. His “and therefore…” is where he loses the plot and starts making baseless claims.

[Richard Vermillion @rivermillion]

@8teAPi Heh

[Richard Vermillion @rivermillion]

@TheDataVet @AsyncAndy @ChristianHeiens The exact phrase “melting pot” was in use by 1910. The concept of America as a nation where individuals of all nations “melted” together was already in use in the late 18th century and throughout the 19th.

[Richard Vermillion @rivermillion]

@RaminNasibov Ultima III

Wizardry and Bard’s Tale are close.

Hunt the Wumpus, but that was pre-1985.

[Richard Vermillion @rivermillion]

@jessi_cata Well, I happen to believe that qualia just are the enacted settlement of a self-body-world model trajectory. Which, under one useful quotient we describe as “seeing red”.

[Richard Vermillion @rivermillion]

@gro_tsen And quartz, of course.

[Richard Vermillion @rivermillion]

@mmjukic @pmarca Because all of the plausible routes that would let the intelligent and moral skirt egalitarianism and take power would also let the stupid or immoral take power, and so the systems that last have built defenses against them.

[Richard Vermillion @rivermillion]

Simulation is only a coherent concept where being and enactment come apart. For a thing constituted by its enactment, “perfect simulation” is not simulation, it’s instantiation.

The argument is about whether selves are constituted by their enactment or not.

[Richard Vermillion @rivermillion]

This is such a messed up, competitive conception of what intelligence is for. Interacting with more intelligent agents makes intelligence more worthwhile. It’s not a zero sum game. Your intelligence is not meaningful because others don’t have it. To the contrary.

[Richard Vermillion @rivermillion]

@GaryMarcus You literally can’t prove or disprove that dogs are conscious, so should we return to the days where we confidently state that they don’t suffer and only “respond to stimuli”?

We shouldn’t demand proof of a narrow conception of consciousness to consider extending care.