195 mentions · 158 chunks · 69 episodes
ChatGPT Atlas's coverage has been dominated by stories about prompt injection and privacy. A prompt injection attack was demonstrated in the first 24 hours. ...
Fortune: An ex-OpenAI researcher's study of a million-word ChatGPT conversation shows how quickly 'AI psychosis' can take hold—and how chatbots can sidestep safety guardrails
Apparently ChatGPT's subscriber counts are stalling. I hadn't realized how much of ChatGPT usage is free users. Only 5% (!!) of users are paying subscribers.[im][in] It...
OpenAI is looking at a Log in with ChatGPT offering. A crucial feature: businesses that use it can subsidize inference for users. This is a classic aggregator play to lock in an ecosystem base...
The Atlantic points out that ChatGPT is a fictional character. But unlike many characters, it was not authored. Instead, it emerged. A very different dynamic. Yet another way that the im...
A few thoughts on ChatGPT allowing adult content. It leans into hyper addiction and engagement even more. At that scale of deployment (beyond some niche for things like Replik...
Stratchery dove into why Walmart decided to integrate with ChatGPT Commerce. To me this implies that the current model of software is stagnant.
...e platform owner's permission? It's a fundamentally different kind of thing. So ChatGPT is not like Windows, it's like Facebook, but turbocharged. Because even on Facebook, it's only valuable if other people and other content is on it. C...
ChatGPT Commerce shines when finding candidates is easy but verifying they match your goals is tedious. LLMs have infinite patience and OK judgment. Two conc...
When ChatGPT Turns Informant "The largely overlooked privacy risks of using AI apps that not only remember your conversations, but are capable of using these to r...
People bought windows (and even iPhones) largely because of 3P apps. But with ChatGPT, 99.999% of experiences are 1P. An 'ecosystem' that is so hyper-concentrated is not resilient.
The new Buy it in ChatGPT feature, powered by Stripe, seems like a fancy iframe. But one where the entity controlling it can easily be tricked. It would be a conflict of inter...
...acting those insights takes work… and it's easy for the system to get it wrong. ChatGPT does this with its dossier, which it's continually expanding, tweaking, based on your most recent chats. But it doesn't share that dossier with you. ...
... Also from Futurism: In Situations Where Most Humans Think You're Being a Jerk, ChatGPT Will Assure You You're Behaving Like an Angel.
I loved this riff on the Pelican+baby meme about ChatGPT and your data: "AI SYSTEM perfect size for put data in to secure! inside very secure and useful data will be useful put data in AI System. Put data i...
...sensitive medical questions invade your privacy by giving data to OpenAI? Using ChatGPT vs using a service that uses the same model via the API are orders of magnitude different risk. In the former, OpenAI explicitly reserves the right t...
...ShadowLeak: A Zero-Click, Service-Side Attack Exfiltrating Sensitive Data Using ChatGPT's Deep Research Agent.. Ars Technica's summary Eito Miyamura: "We got ChatGPT to leak your private email data… …All you need? The victim's email addr...
ChatGPT Sent Me To The ER The title is click-bait-y, but the author actually did have a rare condition that required immediate attention. LLMs have infinite ...
...xporting easy, too. This article compares how the philosophy is the opposite of ChatGPT.
A signpost: an article in the Washington Post telling consumers how to disable ChatGPT from training on your conversations.