TZUTC: -0500   
   MSGID: 1050.consprcy@1:2320/105 2c6a4463   
   PID: Synchronet 3.20a-Linux master/acc19483f Apr 26 202 GCC 12.2.0   
   TID: SBBSecho 3.20-Linux master/acc19483f Apr 26 2024 23:04 GCC 12.2.0   
   BBSID: CAPCITY2   
   CHRS: ASCII 1   
   ChatGPT can remember more about you than ever before  should you be worried?   
      
   Date:   
   Sat, 19 Apr 2025 19:30:00 +0000   
      
   Description:   
   ChatGPTs new memory features promise smarter, more personal responses  but at   
   what cost to privacy, control, and connection?   
      
   FULL STORY   
   ======================================================================   
      
   ChatGPTs memory used to be simple. You told it what to remember, and it   
   listened.    
      
   Since 2024, ChatGPT has had a memory feature that lets users store helpful   
   context. From your tone of voice and writing style to your goals, interests,   
   and ongoing projects. You could go into settings to view, update, or delete   
   these memories. Occasionally, it would note something important on its own.   
   But largely, it remembered what you asked it to. Now, thats changing.    
      
   OpenAI, the company behind ChatGPT, is rolling out a major upgrade to its   
   memory . Beyond the handful of facts you manually saved, ChatGPT will now    
   draw from all of your past conversations to inform future responses by    
   itself.    
      
    According to OpenAI , memory now works in two ways: saved memories, added   
   directly by the user, and insights from chat history, which are the ones that   
   ChatGPT will gather automatically.    
      
   This feature, called long-term or persistent memory, is rolling out to    
   ChatGPT Plus and Pro users. However, at the time of writing, its not    
   available in the UK, EU, Iceland, Liechtenstein, Norway, or Switzerland due    
   to regional regulations.    
      
   The idea here is simple: the more ChatGPT remembers, the more helpful it   
   becomes. Its a big leap for personalization. But its also a good moment to   
   pause and ask what we might be giving up in return.   
      
   A memory that gets personal   
      
   Its easy to see the appeal here. A more personalized experience from ChatGPT   
   means you explain yourself less and get more relevant answers. Its helpful,   
   efficient, and familiar.    
      
   Personalization has always been about memory, says Rohan Sarin, Product   
   Manager at Speechmatics , an AI speech tech company. Knowing someone for   
   longer means you dont need to explain everything to them anymore.    
      
   He gives an example: ask ChatGPT to recommend a pizza place, and it might   
   gently steer you toward something more aligned with your fitness goals  a   
   subtle nudge based on what it knows about you. It's not just following   
   instructions, its reading between the lines.    
      
   Thats how we get close to someone, Sarin says. Its also how we trust them.   
   That emotional resonance is what makes these tools feel so useful  maybe even   
   comforting. But it also raises the risk of emotional dependence. Which,   
   arguably, is the whole point.    
      
   From a product perspective, storage has always been about stickiness, Sarin   
   tells me. It keeps users coming back. With each interaction, the switching   
   cost increases.    
      
   OpenAI doesnt hide this. The company's CEO,. Sam Altman, tweeted that memory   
   enables AI systems that get to know you over your life, and become extremely   
   useful and personalized.    
      
   That usefulness is clear. But so is the risk of depending on them not just to   
   help us, but to know us.   
      
   Does it remember like we do?   
      
   A challenge with long-term memory in AI is its inability to understand    
   context in the same way humans do.    
      
   We instinctively compartmentalize, separating whats private from whats   
   professional, whats important from whats fleeting. ChatGPT may struggle with   
   that sort of context switching.    
      
   Sarin points out that because people use ChatGPT for so many different    
   things, those lines may blur. IRL, we rely on non-verbal cues to prioritize.   
   AI doesnt have those. So memory without context could bring up uncomfortable   
   triggers.    
      
   He gives the example of ChatGPT referencing magic and fantasy in every story   
   or creative suggestion just because you mentioned liking Harry Potter once.   
   Will it draw from past memories even if they're no longer relevant? Our   
   ability to forget is part of how we grow, he says. If AI only reflects who we   
   were, it might limit who we become.    
      
   Without a way to rank, the model may surface things that feel random,   
   outdated, or even inappropriate for the moment.   
      
   Bringing AI memory into the workplace   
      
   Persistent memory could be hugely useful for work. Julian Wiffen, Chief of AI   
   and Data Science at Matillion , a data integration platform with AI built in,   
   sees strong use cases: It could improve continuity for long-term projects,   
   reduce repeated prompts, and offer a more tailored assistant experience," he   
   says.    
      
   But hes also wary. In practice, there are serious nuances that users, and   
   especially companies, need to consider. His biggest concerns here are    
   privacy, control, and data security.    
      
   I often experiment or think out loud in prompts. I wouldnt want that retained    
    or worse, surfaced again in another context, Wiffen says. He also flags    
   risks in technical environments, where fragments of code or sensitive data   
   might carry over between projects, raising IP or compliance concerns. These   
   issues are magnified in regulated industries or collaborative settings.   
      
   Whose memory is it anyway?   
      
   OpenAI stresses that users can still manage memory  delete individual    
   memories that aren't relevant anymore, turn it off entirely, or use the new   
   Temporary Chat button. This now appears at the top of the chat screen for   
   conversations that are not informed by past memories and won't be used to   
   build new ones either.    
      
   However, Wiffen says that might not be enough. What worries me is the lack of   
   fine-grained control and transparency, he says. It's often unclear what the   
   model remembers, how long it retains information, and whether it can be truly   
   forgotten.    
      
   Hes also concerned about compliance with data protection laws, like GDPR:    
   Even well-meaning memory features could accidentally retain sensitive    
   personal data or internal information from projects. And from a security   
   standpoint, persistent memory expands the attack surface. This is likely why   
   the new update hasn't rolled out globally yet.    
      
   Whats the answer? We need clearer guardrails, more transparent memory   
   indicators, and the ability to fully control whats remembered and whats not,"   
   Wiffen explains.   
      
   Not all AI remembers the same   
      
   Other AI tools are taking different approaches to memory. For example, AI   
   assistant Claude doesnt store persistent memory outside your current   
   conversation. That means fewer personalization features, but more control and   
   privacy.    
      
    Perplexity , an AI search engine, doesnt focus on memory at all  it    
   retrieves real-time web information instead. Whereas Replika, AI designed for   
   emotional companionship, goes the other way, storing long-term emotional   
   context to deepen relationships with users.    
      
   So, each system handles memory differently based on its goals. And the more   
   they know about us, the better they fulfill those goals  whether thats    
   helping us write, connect, search, or feel understood.    
      
   The question isnt whether memory is useful; I think it clearly is. The   
   question is whether we want AI to become this good at fulfilling these roles.    
      
   Its easy to say yes because these tools are designed to be helpful,    
   efficient, even indispensable. But that usefulness isnt neutral, its   
   intentional. These systems are built by companies that benefit when we rely    
   on them more.    
      
   You wouldnt willingly give up a second brain that remembers everything about   
   you, possibly better than you do. And thats the point. Thats what the   
   companies behind your favorite AI tools are counting on.   
      
   ======================================================================   
   Link to news story:   
   https://www.techradar.com/computing/artificial-intelligence/chatgpt-can-rememb   
   er-more-about-you-than-ever-before-should-you-be-worried   
      
   $$   
   --- SBBSecho 3.20-Linux   
    * Origin: capitolcityonline.net * Telnet/SSH:2022/HTTP (1:2320/105)   
   SEEN-BY: 105/81 106/201 128/187 129/305 153/7715 154/110 218/700 226/30   
   SEEN-BY: 227/114 229/110 111 114 206 300 307 317 400 426 428 470 664   
   SEEN-BY: 229/700 705 266/512 291/111 320/219 322/757 342/200 396/45   
   SEEN-BY: 460/58 712/848 902/26 2320/0 105 3634/12 5075/35   
   PATH: 2320/105 229/426