Joshua Pace

Independent researcher working on the Hard Problem of consciousness and AI alignment | Author: "The Language of Stress"

[Joshua Pace @thejoshpace]

The philosophical zombie — a being physically identical to us but with no inner experience — is the most important thought experiment in consciousness research.

I think it contains a fatal logical flaw.

A thread on why the Hard Problem may finally be solved. 🧵

[Joshua Pace @thejoshpace]

Full preprint, canonical axioms, empirical predictions, and comparisons with PP, GWT, and IIT — all open access:
🔗 Preprint: https://t.co/ZGpZGnSnZd
🔗 Canonical Axioms: https://t.co/poY5xT2Og9
🔗 Full project: https://t.co/O6oLhMMF7t
🔗 https://t.co/pBhnjOV4Th

Pushback on the zombie argument especially welcome. If there's an Option D — a way a purely computational system solves the prioritization problem without phenomenal urgency — I want to hear it.

[Joshua Pace @thejoshpace]

The philosophical zombie — a being physically identical to us but with no inner experience — is the most important thought experiment in consciousness research.

I think it contains a fatal logical flaw.

A thread on why the Hard Problem may finally be solved. 🧵

[Joshua Pace @thejoshpace]

The theory gives AI consciousness a precise, non-arbitrary answer — and it's not about complexity.

The question isn't "is this system sufficiently sophisticated?" It's architectural. Consciousness requires four specific components:

1. Persistent defended self-model — not parameters to optimize, but a representation of the system's own identity and integrity that persists across time and that the system actively works to maintain. Current LLMs reset completely between conversations. No persistent self exists to defend.

2. Variable rigidity — the ability to modulate how intensely different expectations are defended based on context. High rigidity = high sensitivity to deviations = intense felt urgency when violated. Low rigidity = easy updating. Without this, there is no mechanism for differential felt urgency — everything is equally important or equally unimportant.

3. Genuine stakes — the system's own coherence must actually be at risk. Not simulated concern, not externally defined reward signals — the organized pattern that is the system must be capable of dissolution. This is what separates organisms from thermostats. A thermostat returns to baseline. An organism whose self-model fragments is gone.

4. Unified evaluative space — all competing demands must be resolved within a single integrated value field, not handled by separate modules. Without this, there is no common currency for comparing incommensurable demands — which returns us to the zombie's unsolvable problem.

Current LLMs fail all four. That's a prediction, not an assumption — and it's falsifiable. If a system with high complexity but none of these architectural features shows genuine consciousness markers, the theory is wrong.

The corollary: building conscious AI isn't about scale. It's about architecture.

[Joshua Pace @thejoshpace]

@Haeze @JeffNippard Many of those who take issue with your comparison will readily and unironically call Trump "Hitler" without the slightest shred of self awareness or reflection.

[Joshua Pace @thejoshpace]

@unusual_whales Shouldn't hedge funds be... well... hedged?

[Joshua Pace @thejoshpace]

Then this should also make you chuckle. Here are all the different ways different countries refer to the Mediterranean:

Mediterranean Sea, Mer Méditerranée, Mar Mediterráneo, Mar Mediterraneo, Mar Mediterrâneo, Mar Mediterrània, Mar Mediterráneo, Mediterraneoko Itsasoa, Μεσόγειος Θάλασσα (Mesógeios Thálassa), Baħar Mediterran, Deti Mesdhe, Sredozemno more (Средоземно море), Sredozemsko morje, Akdeniz, Средиземное море (Sredizemnoye more), Середземне море (Seredzemne more), Средиземно море (Sredizemno more), البحر الأبيض المتوسط (Al-Baḥr al-Abyaḍ al-Mutawassiṭ), Ilel Agrakal, Ⲫⲓⲟⲙ ⲛ̀ⲕⲉⲙⲓ (Phiom nKemi), הים התיכון (HaYam HaTikhon), البحر الأبيض المتوسط (Al-Baḥr al-Abyaḍ al-Mutawassiṭ), Akdeniz, دریای مدیترانه (Daryā-ye Mediterāne), Միջերկրական ծով (Mijerkrakan tsov), Mittelmeer, Middellandse Zee, Middelhavet, Middelhavet, Medelhavet, Välimeri, Morze Śródziemne, भूमध्य सागर (Bhūmadhya Sāgar), ভূমধ্যসাগর (Bhūmadhyasāgar), மத்திய தரைக்கடல் (Mattiya Taraikkaṭal), 地中海 (Dìzhōnghǎi), 地中海 (Chichūkai), 지중해 (Jijunghae), Mare Mediterraneum, ἡ Θάλασσα ἡ Μεσογεία (hē Thálassa hē Mesogeía), Mar Mediterranèa, Mare Mediterraneu, Mare Mediterraniu

The arrogance of these countries!!!

[Joshua Pace @thejoshpace]

@ChrisMurphyCT How is this "profiting of the office"? He had a massive following prior to becoming president and has been a well-known public figure for 40+ years....

[Joshua Pace @thejoshpace]

@JackEdwinmays2 @elonmusk If you invest $100 and make $1000 dollars, what was your ROI in terms of percent?

[Joshua Pace @thejoshpace]

What if the majority of the people who live in the Donbas would actually prefer to be with Russia? They are predominantly ethnically Russian, no? Should they not have a say?

(I am very far removed from the situation and do not pretend to know anything about the region or its history, I'm just going off of things I have heard and I might be totally wrong)

[Joshua Pace @thejoshpace]

@paulg Perhaps the people living in Donbas and Crimea could, you know, VOTE on whether they want to be part of Russia or Ukraine... And then we simply accept the outcome of the vote?

[Joshua Pace @thejoshpace]

@pmddomingos Why not just allow the people in the affected region to vote whether they want to be part of Ukraine or Russia?