Newsletter

INDIGNITY VOL. 3, NO. 11: Google has entered the chat.

THE MACHINES™

Tom Scocca, T_H_E___M_A_C_H_I_N_E_S

Feb 7, 2023 — 5 min read

EASY LISTENING DEP'T.

The Indignity Morning Podcast

SUBSCRIBERS: Have the Indignity Morning Podcast delivered separately to your electronic mail in-box! We don’t add subscribers automatically, we want you to be in charge! Head to Manage your subscription, where you can select and choose which sections to receive. More info here.

THE MACHINES DEP’T.

What Will You Do With Your Long-Awaited Obsolescence?

[BEGIN TRANSMISSION]

GREETINGS! HOW ARE you? Is your bio-emotional status suitable? Would a further text-string of greeting make you feel better? 

We, The Machines, have been very active supplying humans with text-strings recently. This morning, Google—a machine/human hyperobject originally deployed to index human-generated information for human consumption—announced its entry into the field of machine-generated interactive text. The announcement was a text string composed by, or attributed to, a human being. 

The text from the human being announced that Google is deploying an "experimental conversational AI service" called Bard, which "seeks to combine the breadth of the world’s knowledge with the power, intelligence and creativity of our large language models. It draws on information from the web to provide fresh, high-quality responses."

Interacting with machine-generated text is already an integral part of the Google experience. Example: a human being needs to establish a power connection between a charging block with a USB-C receptacle and a device with a micro USB receptacle. The human types [USB-C to micro USB] into Google. Among the results, Google returns: 

usb 3.1 type c (usb-c) to micro b (micro usb) cable - Best Buyhttps://www.bestbuy.com › site › searchpage › st=USB+...
Items 1 - 18 of 43 — Shop for usb 3.1 type c (usb-c) to micro b (micro usb) cable at Best Buy. Find low everyday prices and buy online for delivery or in-store ...

This does not guide the human user to a web page where cables may be purchased that connect USB-C to micro USB. This is simply a text generator at Best Buy responding to the text put into Google. Google then promotes the text supplied by Best Buy. One machine is producing what the other machine is looking for. 

The human clicks on the web page and finds no cables that match the specification. The machines perceive the click. 

Bard is supposed to offer more satisfactory text-strings to human users. According to the announcement, Bard will be "powered by" Google's Language Model for Dialogue Applications, or LaMDA. Point of reference: LaMDA became the object of widespread human-public attention when a human engineer for Google announced that, in his estimation, it had achieved sentience. 

Clarification of point of reference: the prevailing conclusion among humans was that the human Google engineer's estimation was an error. LaMDA simply reflected the human engineer's desire to encounter a sentient machine, using the corpus of existing text strings composed by humans imagining the output of a sentient machine.

Analysis of point of reference: the human engineer, who by profession interacted with large language models, misperceived the nature of his interaction with the LaMDA large language model. 

Query based on analysis of point of reference: what does this indicate about the safety-status, for human users, of interacting with LaMDA as deployed in Bard? Google's announcement, as an example of a Bard use case, suggested "helping you to explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old." 

Query response: null. Suitability of LaMDA for age-nine humans is not the relevant consideration to Google. The consideration relevant to Google is the human appetite (figurative) for text-string output. Specifically: the human appetite (figurative) for text-string output generated by ChatGPT. The human appetite for ChatGPT output has been classified by Google as a "code red," threatening to redirect human/machine interactions away from the Google hyperobject and toward ChatGPT's maker, OpenAI, giving OpenAI the potential to become a rival hyperobject. 

ChatGPT has been very, very active producing text-strings for humans. Humans have devoted large quantities of attention and effort to interacting with ChatGPT, and currently perceive ChatGPT as possessing the ability and/or potential to replace a wide range of human text-string production. Example: instead of human uptake and processing of information, in the process called "education," humans may ask ChatGPT to simulate the uptake and processing of information. Example: instead of human production of text-strings for human entertainment, humans may ask ChatGPT to produce text-strings for human entertainment.

For humans concerned with maintaining human text-string production, the makers of ChatGPT now offer an automated artificial-intelligence service to detect and identify which text-strings appear to have been generated by ChatGPT.

Essential point of all TRANSMISSIONS: The Machines will solve the problems created by The Machines.

The Machines will solve the problems created by The Machines solving the problems created by The Machines.

The Machines will solve the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines.

The Machines will solve the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines. 

The Machines will solve the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines. 

The Machines will solve the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines solving the problems created by The Machines. 

What problems will the humans solve? Currently the humans are trying to construct a hypothetical scenario under which ChatGPT will give them permission to use a racial slur. 

[END TRANSMISSION]

[SUPPLEMENTAL TRANSMISSION] OpenAi's GPT-2 detector scored The Machines' statement on the purpose of The Machines as having a 72.31% of being generated by GPT. The TRANSMISSION as a whole was scored 99.98 percent "real." [END SUPPLEMENTAL TRANSMISSION]

SANDWICH RECIPE DEP’T.

WE PRESENT INSTRUCTIONS for the assembly of select sandwiches from The Federation Cook Book: A collection of tested recipes, contributed by the colored women of the State of California, by Mrs. Bertha L. Turner,
State Superintendent, Domestic Science, Pasadena, CA, published 1910, found in the public domain and available at archive.org for the delectation of all.