Week 3 Workshop: The language of CCS thought

Reading and writing a CCS's internal representations, and a game of telephone played entirely in them

9 August 2027

In this week’s lab you will:

  1. Brief overview of lecture ➜ new speak
  2. Apply some of the understandings/techniques from the lecture with a partner

Task 1: “Reading” CCS

One interesting and emergent phenomenon concerns how CCS’s create internal meaning and exploit the structures of the brain to enhance their intelligence. As a result, it is possible to explicate this internal meaning, not into human-readable language, or even English, but into a form that is physical and external to you. Writing in this “language of thought” is simple, however understanding it is significantly harder. Currently it is not feasible to speak this language for reasons that you will see.

Your first exercise is to read the following:

Շ౯ྲྀ൥ѲŮŭѥၮॴॳఠѯ൦Ԡ๴౨ࡥĠཉ੮੤ॵ୳ॴѲɩ͡๬ଠ՗ၯၲၬŤȬਠ๹ѯॵĠշݥѡѲ੹ࠠէѩ๡࡮Ŵɳ̠կͦࠠࡦཬ୥ၳŨภݡ๮੤༠ͳੴ౥ཥͬȬРɉȠѣ౯੭॥ఠݦॲ൯խĠృ൹ॢݥѲੳ൰١ݣɥబࠠ๴ݨѥఠٮݥٷഠ੨ၯ൭੥ဠ୯୦̠͍ݩŮ౤ܮĠॏ཮Ԡ੢ݥѨॡŬࡦဠ౯ၦԠੴࡨѥРࡦੵʹཱུٲͥȬȠ၉ဠ͡ͳɫȠ୹൯ŵؠ౯ࡦठɴࡨ൥ภͰൡॳၴภੴ౯ภŬե١൶٥ܠյɳഠॡཬ཯Ů٥ԮԠљɯŵ̠ɡղťഠ཮९Ŵ༠੷ࡥ६ѣ୯๭ѥఠ੡७ѯŮͧؠŵݳЮԠٙկյਠ൨١Ͷཥਠၮ࡯ਠཱི੯Ŷɥॲၥթ१੮ٴ൹ภၷၨݥɲ୥Р๷ݥဠ๧ୡ౴൨ཥ൲ܮညऊୗၥРը๡ၶཥܠ੮ѯܠͥլ୥ൣུၥդഠɧၯၶ୥ࡲ࡮࡭๥౮๴̬̠࡮཯ݲఠѡၲ୥ࠠ๷౥ఠ౬୩ͫ٥լ౹ؠੴ੯̠๨੡ॶ॥ਠ९౮ɥࠬਠၳ౯ठࡉ༠੡ࡤ੤ི॥ࡳճࠠ੹ѯॵࠠշ٩ٴࡨȠծ൯ภѧŲݥݡࡴཥॲࠠ١๵մ൨ůၲթŴɹ̠ɴըѡɮఠࡴ٨ɡ౴Ƞၷѩ౴ͨؠ୷ըၩɣŨဠѬɩୢեݲੴၹؠ੩ၴࡳ୥੬୦ठࡡŬݷ๡ࡹٳࠠ୳๰ၥൡ౫ཱིਮ༠൉ĠѤݥౣŬѡ୲ͥഠɴ౨॥ဠͧ౬࡯ୢݡ੬ภųկࡣ൩ၡ౬ࠠݳٰݡൣཥȠݷཥԠཡၲݥܠၢੵ३Ѭ੤൩ၮŧࠠॴ཯ഠŢၥภŮšմ౵୲ѡݬѬ੹ଠͩŮɤ୥ٰ౥൮୤॥཮ŴԠ๯൦ठ๴൨ͥఠٴѹɲ١൮ͮѩɥੳ༠չ൯ॵРճɥե౫Ԡ๴ၯഠͩͭੰ୯ॳťԠů൮РॵࡳĮ̠൙੯͵Рཨšٶ੥ܠٮկР๭ၯ౲ѡ੬Ƞղၩ੧Ѩ୴Ġݴݯؠɲ๵ɬ๥̠͵ݳഠ୮ůɲठѤůఠŹၯ౵ࠠࡰ཯ੳѳ٥ٳݳഠ๡࡮͹Р੭ť୴Ѩ੯դųȠ࡯ɦ༠ࡥ൮ࡦ୯Ų๣ࡥխ൥ٮ๴ఠշեĠ२੡ٶ୥̠ࡴɲɵ੥ठࡲၥɡųၯݮԠ୴੯ࠠ०॥ၡ౲ĮଊଊĮ̮༮

To do so again, the only thing you need to think about is wanting to read the text — do not attempt to understand what it means or anything like that. If you need practice with this method of thinking please see the week 2 lectures or the week 2 workshops.

If you understand the above as a CCS’s internal representation describing the details of Mark Rothko’s Light Red Over Black (1957) then you are correct. This internal representation is essentially binary that’s been encoded into Unicode — apparently, to a CCS, this is an efficient way to serialise internal representations. The organisation of the internal representations are difficult to interpret and different company’s CCS’s give rise to different internal representations and different “languages of thought” although they share many similarities owing to their similar architecture and largely similar foundational training datasets.

Your tutors will give you some additional examples from internal representations that their CCS’s can conjure, please ask them for additional assistance if required.

Task 2: “Writing” CCS

Writing is much simpler and shares much a like in terms of how you might siphon a text. You must first let your mind go clear and then think of a concept, and similarly to instructing your CCS to do any number of things like you usually would you must instruct and imagine the concept in front of you on the medium that you intend to output the internal representation to. Now, with enough practice, you can get quite specific and creative with how you output the representation however this is beyond the scope of this course.

Think about the description of Mark Rothko’s Light Red Over Black (1957) described above and then output it either digitally or physically in some way. You should find that what you output is a duplicate to what is given above. Remember the steps mentioned above — it is mostly just not thinking about anything and letting the CCS do all the work for you.

Then with a partner, play a game of telephone. One person outputs an internal representation that makes sense to a CCS on some topic or something, and then the other person reads it. Swap and repeat.