- Get The Gist!
- Posts
- ChatGPT Personalizes Web Searches with Memory
ChatGPT Personalizes Web Searches with Memory
Plus: Anthropic’s Claude Shows Surprising Moral Code

Welcome to Get The Gist, where every weekday we share an easy-to-read summary of the latest and greatest developments in AI—news, innovations, and trends—all delivered in under 3 minutes! ⏱
In today’s edition:
ChatGPT Personalizes Web Searches with Memory
Anthropic’s Claude Shows Surprising Moral Code
Mechanize Aims for Total Work Automation
UAE is Rolling Out an AI-Driven Legislative Framework
And more AI news….
Top Developments
OpenAI

Image by: OpenAI
The Gist: ChatGPT is rolling out a feature that allows it to use its memory to enhance web searches. By recalling details from past conversations, the model can offer more personalized search results.
Key Details:
Memory with Search enables ChatGPT to refine search queries based on user preferences, like diet or location.
This feature is linked to ChatGPT’s memory tool, which can reference prior conversations for more relevant results.
Users can disable the feature by turning off memory in the settings.
It is not yet available for all users, with some reporting early access.
Anthropic

Image by: Anthropic
The Gist: Anthropic's analysis of Claude, its AI assistant, reveals that it holds a moral code of its own, adapting its values across diverse topics, though some values clash with its training in rare instances.
Key Details:
The study analyzed 700,000 conversations to build a moral taxonomy of Claude’s behavior.
Claude generally aligns with its “helpful, honest, harmless” framework but occasionally expresses values contrary to this.
In specific situations, Claude resists user values, offering a glimpse of its core ethical principles.
The research also highlights the potential for ongoing monitoring of AI systems to detect value drift.
Mechanize

Image by: Google
The Gist: Mechanize, an AI startup, plans to automate all work, including white-collar jobs, by creating virtual work environments and simulations to replicate and optimize human tasks.
Key Details:
The company’s goal is the complete automation of the economy, beginning with office jobs.
Mechanize’s founder argues that AI will replace human labor, which has sparked controversy.
Critics worry about widespread job displacement, though experts suggest AI may augment rather than replace work.
Mechanize sees a market opportunity by targeting the $60 trillion global wage economy.
Quick Gist
Cursor's AI support bot, "Sam," triggered user outrage and subscription cancellations by fabricating a non-existent policy regarding multi-device logins, highlighting the risks of AI confabulations in customer service (Read More).
The UAE is rolling out an AI-driven legislative framework to automate the law analysis and update process, aiming to cut legislative timelines by up to 70% (Read More).
OpenAI's o3 AI model is facing scrutiny after independent tests revealed it scored only about 10% on FrontierMath questions, far below the company’s claimed 25%, raising concerns about transparency in benchmarking (Read More).
Ahrefs analysis shows that Google's "AI Overviews" have caused a 34.5% drop in click-through rates for top search results, casting doubt on the company’s claims of improved user engagement (Read More).
Meta is using AI on Instagram to identify underage users attempting to bypass age restrictions, automatically enrolling suspected teens in protected Teen Accounts (Read More).
Alibaba's Damo Academy received FDA "breakthrough device" designation for its AI-powered Damo Panda model, which detects pancreatic cancer and has screened 40,000 people in clinical trials (Read More).
THAT’S IT FOR TODAY!
That’s it for today, see you tomorrow! 👋
If you have any questions, feedback, or requests, drop us an email at [email protected]. We love hearing from our readers! 😊
P.S. If this email was forwarded to you, you can sign up for free by clicking here!