@AnthropicAI@OpenAI
So after seeing the new memory update for ChatGPT and thinking about this, this popped into my head and I wanted to share it.
Right now both Claude and now with the new memory update ChatGPT will generate memory based on past conversations.
What I'd love to see: the AI periodically reviews its own memory, flags things that are likely outdated, and asks the user before doing anything.
"Hey, I noted you were apartment hunting in early 2024 β is that still relevant or can I clear it?"
Key details to make it work:
- Time-triggered: only surfaces memories older than X months with no recent references
- Life-event sensitive: relationships, jobs, living situations get flagged before deletion, not auto-purged
- Batched: one clean monthly or yearly summary, not constant popups
Claude and ChatGPT already lets you manually view and manage memory which is great. Proactive surfacing of stale stuff feels like a natural next step. Keeps the user in control without requiring them to babysit it manually.
It's a mix of both. There will be a lot of automation, that's for sure. Focusing more on what matters and delegating repetitive tasks to AI is already a win. The usefulness of these AIs is highly subjective. I use it every day for Q&A, work, and learning new stuff. Those who use AI blindly to reduce costs are driven purely by profit. AI and robots are a double-edged sword. On one side they will take over jobs, but on the other, since companies can produce stuff cheaper, they will also sell it cheaper.
I donβt think they do. The reason is that many parents never taught their kids how to talk about these things or how important communication is. On top of that, the internet twisted every word into βdonβt say that, it might hurt his/her feelings.β If you communicate an opinion positive or negative without the intent to hurt someone, then it shouldnβt automatically be treated as an attack. And if someoneβs opinion hurts your feelings, then deal with it. Life doesnβt care.
I think people fear AI because they see it as a replacement instead of a tool. In the right hands, it boosts productivity, creativity, learning, and problem-solving. Refusing to adapt to it now feels like refusing to learn the internet back then. The people who learn to use it well will simply move faster. I also think a lot of this fear comes from companies using AI to cut costs, even though that was never its original purpose. Iβm sure many people will disagree with me, and I welcome the discussion. I love talking about this topic because I genuinely believe AI is part of the future weβre heading toward.
I agree that the way some companies train and use AI is questionable. But I also think thereβs nuance here.
Humans learn from other peopleβs work too. Artists study other artists, programmers read other code, musicians learn from existing music. Learning patterns from existing work isnβt automatically theft.
The bigger issue is consent and compensation. I think there should be an open standard where creators can choose whether their work can be used for AI training, ideally with attribution or compensation models attached.
That would reduce aggressive scraping while still allowing AI to develop in a more ethical way.
@ARl421@cyber_razz Just because management misuses AI and replaces humans with it when it was not designed for that is not AI's fault but the fault of those who just want more money. You should hate those companies that replace humans with AI rather than enhancing their productivity.
@fenceprobably@cyber_razz Itβs a force multiplier. It was never meant to replace humans. Just because you donβt use it, it doesnβt mean itβs worthless. The AI is the same as a calculator or a computer. Both made humans do stuff faster.
@NicolasLDavout@SMB_Punt@DanielSmidstrup@nikhilbhima While I understand your reasoning, Gemini and ChatGPT currently offer stronger consumer products. OpenAI gets around 60% (in 2025) of its revenue from consumers, while Mistral is mainly focused on enterprise and government use, which leaves its consumer experience behind.
Yo Victoria, take your pretty face, log off the victim card, and actually do some research before opening your mouth.
Rachel Zegler spent years shitting on the original Snow White called the Prince a βstalker,β said the story was weird and outdated, turned it into a woke girlboss lecture, and openly mocked the fans who loved the classic. Then the movie bombed hard.
Thatβs not βracial abuse.β Thatβs audiences rejecting someone who spent years insulting the source material and the people who grew up with it. Actions have consequences. Stop pretending criticism = harassment.
NEW: Former Tennessee police officer who went wild with six officers on the job has settled her lawsuit for $500,000.
Maegan Hall says she was "s*xually groomed" by male police officers & sued the city of La Vergne.
Hall allegedly hooked up with multiple colleagues, performed oral s*x on two officers on the job, had a threesome with an officer and his wife and much more.
Hall blamed her s*xual actions on her troubled marriage and the police department.
"Maegan wasnβt looked at like a rookie cop to be trained and promoted. She was looked at like a piece of meat to be sexualized and exploited," her lawyer Wesley Clark said.
"The La Vergne Board of Mayor and Aldermen voted to authorize the mayor to sign a settlement agreement between the City of La Vergne and former police officer Maegan Hall," a statement from the city said.
"The Cityβs insurance provider will pay the sum of $500,000 to Hall as a gross settlement, which includes court costs, attorneys fees, and expenses."
Tesla FSD brakes for an invisible van π
Just two guys casually talking while being driven around by FSD.. we thought: HUH?!
Why would it stop, ah ofcourse FSD already saw what two Human Pilots failed to recognize.
A wild invisible van appeared
FSD didnβt even blink & said: hold my beer
@robotinreallife
Today weβre giving an update on ramping F.03 production at BotQ
In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour
We will manufacture 55 humanoid robots this week
Scam Altman didnβt tell the OpenAI board that he OWNED the OpenAI Startup Fund.
Altman lied in congressional testimony that he didnβt have financial gain from OpenAI.
A completely local agent that lives right inside your browser.
Powered by Gemma 4 E2B and WebGPU, it uses native tool calling to:
π Search browsing history
π Read and summarize pages
π Manage tabs
100% local. No servers needed!