Fear of a Black Pope!

Somebodies think Sundar Pichai should be fired from his job as CEO of Google because of Gemini? Seriously? & so Max Read wins the internet today…

Subscribe now


Max Read wins the internet today!

Max Read wins the internet with this observation: We have finally found something that convinces Effective Accelerationists worldwide that the deployment of AI models needs to be massively slowed down! Simply have one draw a picture of a Black Pope!

And that really is, truly, “objectively, extremely funny!”

People desperately want GPT models to be computer demi-gods, one hair’s width away from truly Turing-class artificial intelligences, capable of bringing them any desired piece of true knowledge about the world to you in an instant, and capable of drawing coherent and readable text or semi-professional images that is better and faster than you could produce to whatever specifications you wish to set.

But that is not what they are. They are weird and buggy natural-language interfaces to weird and buggy database summarization engines accessing really weirdly constructed and unreliable databases.

I did not think that Chat-GPT needed to be shut down or that Sam Altman needed to be fired from OpenAI when Chat-GPT told me that my podcast cohost Noah Smith was not a person but rather a ChatBot maintained by DeLong Technology Systems.

So what is the offense that requires that Gemini be classified as a “lunatic”, its development be paused, its programmers sent to the reëducation camps, and the presiding CEO fired—just because Gemini has gone off the rails with this particular combination of heavy-handed RLHF and hallucination?

It cannot be that it is producing unreliable information, can it? I mean, we already knew that if you ask a ChatBot a question the answer to which you do not know, and it gives you an answer, you still do not know the answer to your question.

At some level, the source of the panic is the demonstration that Gemini does not “know” that, often, people regard race or gender as absolutely essential. It is not essential to humans the a pope be in the city as opposed to the countryside, be smiling as opposed to frowning, be middle-aged as opposed to elderly. But it is—for lots and lots and lots of people—very important that the class of admissible images of popes contain only white men.

And, I suppose, it is also a lot of people who have not yet grasped that there is no such thing as a “mind” back there in Gemini producing the answers.

Share


Max Read on the latest internet panic:

Max Read: Google made an A.I. so woke it drove men mad: ‘Someone asked [Google’s Gemini] to draw a picture of a pope:

“Google’s AI Is an Anti-White Lunatic,” wrote the Thielite investor Mike Solana…. Nate Silver accused it of having “the politics of the median member of the San Francisco Board of Supervisors.” “Every single person who worked on this should take a long hard look in the mirror,” tweeted the e/acc influencer @Frantastic_7…. “They need to shut Gemini down,” Silver tweeted…. The normally reserved business blogger Ben Thompson went absolutely nuclear and called for Alphabet C.E.O. Sundar Pichai to resign….

It is, objectively, extremely funny that Google created an A.I. so woke and so stupid that it drew pictures of diverse Nazis, and even funnier that the woke A.I.’s black pope drove a bunch of MBAs who call themselves “accelerationists” so insane they expressed concern about releasing A.I. models too quickly.… Google… turned the racism dial down too far and… generated another, admittedly much funnier and weirder kind of bias…. The equivocation over Hitler and Elon Musk is… the ironic outcome of guardrails designed to render the program as anodyne and inoffensive as possible.…

It’s hard for me to imagine caring about—let alone getting mad at!—a computer that generates text equivocating between Pol Pot and Martha Stewart because I would never ask a computer to compare the two of them. I have not yet outsourced my research abilities or critical faculties or moral compass to the probabilistic text generator…. Many people believe that chatbots should generate text that reflects their values…. [But] the original question here [should be]: What is Gemini for, exactly?… obtaining true or accurate information?… making text or images to any parameter the user desires?… The answer to both of these questions is “no”…. But… many people inside and outside the software industry are deeply emotionally and often financially invested in the answer to both questions being “yes.”…

Actual… uses for chatbots (summarizing… and… programming tasks, so far?) diverge… from what… enthusiasts want the chatbots to be (computer demi-gods)…. Google and OpenAI and their peers and boosters have marketed A.I. chatbots as all of the above—do-anything miracle tools—but these models… as John Herrman writes… [are] “profoundly strange pieces of software that synthesize averaged-out content from troves of existing media at the behest of users who want and expect countless different things…”

Read Max
Google made an A.I. so woke it drove men mad
Greetings from Read Max HQ! In an effort to play around with format and stave off looming burnout, this week’s newsletter features three shorter blurbs about recent articles or news events. Below you’ll find some thoughts on: How to think about and understand the extremely funny Google Gemini black-pope controversy…
Read more

Share Brad DeLong’s Grasping Reality

Leave a comment

Subscribe now