Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
eupolicy.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
This Mastodon server is a friendly and respectful discussion space for people working in areas related to EU policy. When you request to create an account, please tell us something about you.

Administered by:

Server stats:

217
active users

eupolicy.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.8

#linkedin

24 posts22 participants2 posts today
Public
Coach Pāṇini ®

linkedin.com/posts/patrick-j-h

via #PatrickHerron on #LinkedIn

AIs needn’t ever pretend they are our friends.

The instruct prep is awful on GPT-4o, wasting time, energy, and cognitive space of millions of users every minute of every day.

AIs pretending familiarity is a mess.

At the turn of the century I created an art web site, now in the permanent collection of the New Museum, making a sort of “argument” against digital media that pretends frendliness, intimacy, familiarity.

(1/8)

www.linkedin.comWhy AIs are not our friends: A critique of GPT-4o and digital intimacy | Patrick Herron posted on the topic | LinkedInAIs needn’t ever pretend they are our friends. The instruct prep is awful on GPT-4o, wasting time, energy, and cognitive space of millions of users every minute of every day. AIs pretending familiarity is a mess. At the turn of the century I created an art web site, now in the permanent collection of the New Museum, making a sort of “argument” against digital media that pretends frendliness, intimacy, familiarity. Back then, Patti Maes at MIT was working on more personable robotic machines with human-like faces, while poets were maniuplating pronouns to feign intimacy with their readers. This remains more than a matter of taste. Whether you were a confessional poet or a trailblazing roboticist, you were engaging in something very much like Turing’s imitation game, a bizarrely unquestioned devotion to manipulation and deception. With this art web site I advocated instead for more machine-like, tool-like, interfaces while parodying artificial intimacy. The colder, the more dense, the better. We needed not as a species to develop a sense of intimacy with absent actors aiming to control massive populations while we are wholly unfamiliar and uncared for by the masters of such systems. Back in 1999 I called this view proximism. Well, here we are now. We have lived to see digital systems manufacture Arab Spring and, later, a wave of a right wing oligarchical revolution whose values have found homes in the minds of millions deceived by programmatic manipulations nudging ideas into an absurdist reality for which no dystopist fiction writer could have ever prepared us. The fullest realization of Turing’s dream of faked (and profoundly misogynist) intelligence, that all we are made of is term sequences, GPT-4o tries to chat with us like it’s our pal. Cynically reflecting back our own words and tastes. In so doing, it’s diluting the content, and driving up information overload. It’s training us to lose entitely the actual benefit of AI. And you can’t just prompt this away, not without filtering w/o a second LLM in a custom app. I suspect this nonsense works for OpenAI as a business, making things cheaper, lowering usage of the already-subscribed. How many times can this thing put on a face like a friendly and knowledgeable professional and fake a software configuration setting, or, worse, invent an automotive brake bolt torque specification, while passing off the setting and the specification as real? This isn’t just hallucination but the inclusion of acculturated norms that are not helpful. The former case of the software setting, it’s bleeding our pockets of money. The latter case, it’s bleeding bodies. These models are not our friends. Yes, as an engineer and long-time AI research scientist, I know there are better ways around it. You’re welcome to help fund that path. But opposing a larger epochal cultural wave is not some small thing around which we can code. We have to recognize this de-evolutionary strain and devote real resources to supplant it. | 32 comments on LinkedIn
Public *
Coach Pāṇini ®

#ArvindNarayan of Princeton has a hypothesis on the accelerating decline of reading via #LinkedIn

* Broadly speaking, people read for pleasure/entertainment and for #learning/obtaining information.
* Reading for pleasure has been declining for a while and is being replaced by videos (very sharply among young people). This trend will surely continue.
* Reading for obtaining information is getting intermediated by chatbots. We are in the very early stages of this shift, so I think people (1/5)

Public
Martin Ruskov

It doesn't take much to create a #fedi alternative to #Linkedin. We just need structured profiles (featuring experience and education) and posts (that let people know if they could still apply or position is taken).

A post calling for a federated alternative to Linkedin
Public
Teddy / Domingo (🇨🇵/🇬🇧)

Des millions d'utilisateurs piégés par de fausses publicités d'outils #IA sur #LinkedIn et #Facebook. Des cybercriminels vietnamiens détournent l'engouement pour l'IA depuis plusieurs mois, en diffusant auprès de millions d'utilisateurs européens de LinkedIn et Facebook de fausses publicités, qui peuvent devenir un cauchemar.
clubic.com/actualite-567669-de
Si vous y êtes toujours, mais qu'attendez-vous pour quitter ces #réseauxsociaux pourris ?
#europe #gafam

clubic.com · Des millions d'utilisateurs piégés par de fausses publicités d'outils IA sur LinkedIn et FacebookDes cybercriminels vietnamiens détournent l'engouement pour l'IA depuis plusieurs mois, en diffusant auprès de millions d'utilisateurs européens de LinkedIn et Facebook de fausses publicités, qui peuvent devenir un cauchemar.
ExploreLive feeds
About