Technologies

Gen AI Chatbots Are Starting to Remember You. Should You Let Them?

An AI model’s long memory can offer a better experience — or a worse one. Good thing you can turn it off.

Until recently, generative AI chatbots didn’t have the best memories: You tell it something and, when you come back later, you start again with a blank slate. Not anymore.

OpenAI started testing a stronger memory in ChatGPT last year and rolled out improvements this month. Grok, the flagship tool of Elon Musk’s xAI, also just got a better memory.

It took significant improvements in math and technology to get here but the real-world benefits seem pretty simple: You can get more consistent and personalized results without having to repeat yourself.

«If it’s able to incorporate every chat I’ve had before, it does not need me to provide all that information the next time,» said Shashank Srivastava, assistant professor of computer science at the University of North Carolina at Chapel Hill.

Those longer memories can help with solving some frustrations with chatbots but they also pose some new challenges. As with when you talk to a person, what you said yesterday might influence your interactions today.

Here’s a look at how the bots came to have better memories and what it means for you.

Improving an AI model’s memory

For starters, it isn’t quite a «memory.» Mostly, these tools work by incorporating past conversations alongside your latest query. «In effect, it’s as simple as if you just took all your past conversations and combined them into one large prompt,» said Aditya Grover, assistant professor of computer science at UCLA.

Those large prompts are now possible because the latest AI models have significantly larger «context windows» than their predecessors. The context window is, essentially, how much text a model can consider at once, measured in tokens. A token might be a word or part of a word (OpenAI offers one token as three-quarters of a word as a rule of thumb).

Early large language models had context windows of 4,000 or 8,000 tokens — a few thousand words. A few years ago, if you asked ChatGPT something, it could consider roughly as much text as is in this recent CNET cover story on smart thermostats. Google’s Gemini 2.0 Flash now has a context window of a million tokens. That’s a bit longer than Leo Tolstoy’s epic novel War and Peace. Those improvements are driven by some technical advances in how LLMs work, creating faster ways to generate connections between words, Srivastava said.

Other techniques can also boost a model’s memory and ability to answer a question. One is retrieval-augmented generation, in which the model can run a search or otherwise pull up documents as needed to answer a question, without always keeping all of that information in the context window. Instead of having a massive amount of information available at all times, it just needs to know how to find the right resource, like a researcher perusing a library’s card catalog.

Read more: AI Essentials: 27 Ways to Make Gen AI Work for You, According to Our Experts

Why context matters for a chatbot

The more an LLM knows about you from its past interactions with you, the better suited to your needs its answers will be. That’s the goal of having a chatbot that can remember your old conversations.

For example, if you ask an LLM with no memory of you what the weather is, it’ll probably follow up first by asking where you are. One that can remember past conversations, however, might know that you often ask it for advice about restaurants or other things in San Francisco, for example, and assume that’s your location. «It’s more user-friendly if the system knows more about you,» Grover said.

A chatbot with a longer memory can provide you with more specific answers. If you ask it to suggest a gift for a family member’s birthday and tell it some details about that family member, it won’t need as much context when you ask again next year. «That would mean smoother conversations because you don’t need to repeat yourself,» Srivatsava said.

A long memory, however, can have its downsides.

You can (and maybe should) tell AI to forget

Having a chatbot recommend a gift poses a conundrum that’s all too common in human memories: You told your aunt you liked airplanes when you were 12 years old, and decades later you still get airplane-themed gifts from her. An LLM that remembers things about you could bias itself too much toward something you told it before.

«There’s definitely that possibility that you can lose your control and that this personalization could haunt you,» Srivastava said. «Instead of getting an unbiased, fresh perspective, its judgment might always be colored by previous interactions.»

LLMs typically allow you to tell them to forget certain things or to exclude some conversations from their memory.

You may also deal with things you don’t want an AI model to remember. If you have private or sensitive information you’re communicating with an LLM (and you should think twice about doing so at all), you probably want to turn off the memory function for those interactions.

Read the guidance on the tool you’re using to be sure you know what it’s remembering, how to turn it on and off and how to delete items from its memory.

Grover said this is an area where gen AI developers should be transparent and offer clear commands in the user interface. «I think they need to be providing more controls that are visible to the user, when to turn it on, when to turn it off,» he said. «Give a sense of urgency for the user base so they don’t get locked into defaults that are hard to find.»

How to turn off gen AI memory features

Here’s how to manage memory features in some common gen AI tools.

ChatGPT

OpenAI has a couple types of memory in its models. One is called «reference saved memories» and it stores details that you specifically ask ChatGPT to save, like your name or dietary preferences. Another, «reference chat history,» remembers information from past conversations (but not everything).

To turn off either of these features, you can go to Settings and Personalization and toggle the items off.

You can ask ChatGPT what it remembers about you and ask it to forget something it has remembered. To completely delete this information, you can delete the saved memories in Settings and the chat where you saved that information.

Gemini

Google’s Gemini model can remember things you’ve discussed or summarize past conversations.

To modify or delete these memories, or to turn off the feature entirely, you can go into your Gemini Apps Activity menu.

Grok

Elon Musk’s xAI announced memory features in Grok this month and they’re turned on by default.

You can turn them off under Settings and Data Controls. The specific setting is different between Grok.com, where it’s «Personalize Grok with your conversation history,» and on the Android and iOS apps, where it’s «Personalize with memories.»

Technologies

Today’s Wordle Hints, Answer and Help for Feb. 1, #1688

Here are hints and the answer for today’s Wordle for Sunday, Feb. 1, No. 1,688.

Looking for the most recent Wordle answer? Click here for today’s Wordle hints, as well as our daily answers and hints for The New York Times Mini Crossword, Connections, Connections: Sports Edition and Strands puzzles.

Today’s Wordle puzzle is a tough one, with some unusual letters. If you need a new starter word, check out our list of which letters show up the most in English words. If you need hints and the answer, read on.

Today’s Wordle hints

Before we show you today’s Wordle answer, we’ll give you some hints. If you don’t want a spoiler, look away now.

Wordle hint No. 1: Repeats

Today’s Wordle answer has no repeated letters.

Wordle hint No. 2: Vowels

Today’s Wordle answer has one vowel and one sometimes vowel.

Wordle hint No. 3: First letter

Today’s Wordle answer begins with S.

Wordle hint No. 4: Last letter

Today’s Wordle answer ends with Y.

Wordle hint No. 5: Meaning

Today’s Wordle answer means full of prickles.

TODAY’S WORDLE ANSWER

Today’s Wordle answer is SPINY.

Yesterday’s Wordle answer

Yesterday’s Wordle answer, Jan. 31, No. 1,687, was ALLOT.

Recent Wordle answers

Jan. 27, No. 1,683: DUSKY

Jan. 28, No. 1,684: CRUEL

Jan. 29, No. 1,685: FLAKY

Jan. 30, No. 1,686: JUMBO

Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.

Technologies

Today’s NYT Strands Hints, Answers and Help for Feb. 1 #700

Here are hints and answers for the NYT Strands puzzle for Sunday, Feb. 1, No. 700.

Looking for the most recent Strands answer? Click here for our daily Strands hints, as well as our daily answers and hints for The New York Times Mini Crossword, Wordle, Connections and Connections: Sports Edition puzzles.

Today’s NYT Strands puzzle is a bit of a challenge. Some of the answers are difficult to unscramble, and a couple are kind of long, so if you need hints and answers, read on.

I go into depth about the rules for Strands in this story.

If you’re looking for today’s Wordle, Connections and Mini Crossword answers, you can visit CNET’s NYT puzzle hints page.

Read more: NYT Connections Turns 1: These Are the 5 Toughest Puzzles So Far

Hint for today’s Strands puzzle

Today’s Strands theme is: It’s a gift.

If that doesn’t help you, here’s a clue: For me, really?

Clue words to unlock in-game hints

Your goal is to find hidden words that fit the puzzle’s theme. If you’re stuck, find any words you can. Every time you find three words of four letters or more, Strands will reveal one of the theme words. These are the words I used to get those hints but any words of four or more letters that you find will work:

BONE, GONE, BONNET, NOTE, PRIDE, RING, TING, SENT, RENT, WARD, DRAW, SEEN, SEER, TORE, RANT, TRYING, DONATE, SIRE

Answers for today’s Strands puzzle

These are the answers that tie into the theme. The goal of the puzzle is to find them all, including the spangram, a theme word that reaches from one side of the puzzle to the other. When you have all of them (I originally thought there were always eight but learned that the number can vary), every letter on the board will be used. Here are the nonspangram answers:

AWARD, BONUS, GRANT, PRESENT, DONATION, OFFERING

Today’s Strands spangram

Today’s Strands spangram is GENEROSITY. To find it, start with the G that’s three letters to the right on the top row, and wind down.

Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.

Technologies

I Tested the New AirTag and Found That Apple More Than Doubled Its Range

Review: Apple’s «AirTag 2» gets better at its job. The familiar white and chrome disc is a little louder, higher-pitched and easier to find.

While nearly everyone I know has been freezing and shoveling snow, I spent the past few days under San Francisco’s sunny skies, hiding and finding Apple’s new AirTag. I’m happy to report that it succeeded at its main job: I was able to locate it every time. The second-generation Bluetooth tracker looks and acts identical to the original AirTag, which debuted in 2021, but internal upgrades make it even easier to find.

Over the past five years, the AirTag has become a popular way to track your luggage, keys, car, bike, pets (though you really shouldn’t clip one to Mr. Cupcakes) and nearly anything else you can attach the tiny white and silver disc to. The AirTag also sparked a conversation about privacy, specifically around Apple’s built-in safeguards that prevent it from being used to track someone unwillingly. Apple later expanded those features to protect both iPhone and Android phone owners.

Do we really need a new AirTag? No. But its new features are nice.

This isn’t your typical yearly tech upgrade, like a new phone with a faster processor. The first AirTag is already great. Yeah, it has shortcomings, like not having a hole for a key ring, which forces you to buy a holder to attach it to things. But its best feature is Apple’s Find My network, an encrypted, invisible service connecting over 1 billion devices, including iPhones, Macs and trackers. The AirTag is your key to the Find My kingdom.

In my time testing the second-gen AirTag, I discovered that the latest iteration is also great. It has shortcomings like that missing key ring hole, but the Find My network is still the star. Apple improved the tracker’s ability to be found. The chime is higher-pitched and louder.

When trying to locate it in the Find My app, Precision Finding picked up the new AirTag over twice as far away as it did the old AirTag. The AirTag 2, as we’ve nicknamed it, also supports Precision Finding on my Apple Watch, though setting it up isn’t straightforward.

The second-gen AirTag is on sale now: $29 for one and $99 for a four-pack. And just like the OG version, if you order online directly from Apple, you can get it engraved.

AirTag 1 vs. AirTag 2

When I first unboxed the second-gen AirTag, I was surprised to see that it had the same bulbous Mento design as the original. This allows the new AirTag to be used with the gazillion accessories people already have. But there are differences. The white shell’s resin is now made of 85% recycled plastic. I’m curious to see whether the white casing scuffs as easily as the original.

(That fine print etched on the back? That’s how you can identify a second-gen AirTag from an original: The new one has its text printed in ALL CAPS — except «AirTag,» because Apple presumably wants to keep the intercap.)

If you have the original AirTag, there’s no reason to get rid of it and buy the new version. I can see swapping out an existing AirTag with a second-gen one if you frequently use Precision Finding for important items like keys and everyday bags — and especially if you want to access that feature on your Apple Watch.

If you can choose between the second-gen AirTag and the first-gen model at a discount, I’d go with the discounted original for most items. The first AirTag model does nearly everything the new one can do.

The new AirTag can be found farther away

The new tracker has upgraded Bluetooth connectivity, making it more discoverable to nearby iPhones, which can piggyback its location information anonymously back to its owner. A second-generation Ultra Wideband chip gives the new AirTag a longer range for Precision Finding.

CNET’s Social Media Producer Faith Chihil and I made the trek to Salesforce Park, a green space atop a San Francisco bus station, to run an AirTag test. I gave Faith my old AirTag that I use for traveling, aptly named Patrick’s Luggage, and the new one, named New AirTag. I then walked a few hundred feet away and opened the Find My app on my iPhone 17 Pro Max. I looked for the New AirTag first. I hit the Find button and then got to walking. My iPhone started picking it up 110 feet away. The green «you’re going the right way» screen with an arrow appeared 85 feet away from the new AirTag.

I repeated the test with the Patrick’s Luggage AirTag, and my iPhone started picking it up 42 feet away, and the green screen appeared 37 feet away. In this simple test, the second-gen AirTag was findable at over twice the distance away as my first-gen Apple tracker.

I should note that there are a lot of factors that come into play when trying to find your AirTag. Our range testing was done around lunchtime, so there were a lot of people with iPhones walking by to pick up the tag’s initial location so that my phone could use the Find My network to zero in on it. Also, we were outdoors along a walking path, so we didn’t have walls, rugs or furniture to contend with for the Precision Finding.

The new AirTag supports Precision Finding on the Apple Watch

The second-gen Ultra Wideband chip on the new AirTag also supports Precision Finding for the first time on certain Apple Watch models running WatchOS 26.2.1 or later:

Apple Watch Series 9
Apple Watch Series 10
Apple Watch Series 11
Apple Watch Ultra 2
Apple Watch Ultra 3

But getting it set up isn’t straightforward.

To use Precision Finding on the iPhone, you go to the Items tab in the Find My app and select your AirTag. But when I went to the Find Items app on the watch, I could select the new AirTag, but there wasn’t a Precision Finding option.

A quick online search later, I found an Apple Support page that explained how the Apple Watch uses the Control Center to do it. I had to add a new Find Items button, called Find AirTag, to the Control Center and then tap it to put my watch into Precision Finding mode. Indoors, it found my AirTag from 65 feet away. It might go farther, but our office isn’t that big.

My Apple Watch doesn’t have cellular data, so when I was indoors on Wi-Fi, it worked fantastic. When I was outdoors, connected to my iPhone, it worked mostly fine, but at times it took longer to update my AirTag’s location.

The new AirTag is louder

If you’ve ever used the AirTag’s Play Sound feature through the iPhone’s Find My app, you know that Apple «I’m over here» chime. The new AirTag plays the same tune but is higher-pitched and louder, making it easier to find.

I tested the second-gen AirTag’s chime against the first one using an iPhone audio meter app (not the most scientific, but I wanted to visualize the difference). The original AirTag’s chime peaked at 67.3 dBA, while the new AirTag’s hit 77.5 dBA, more than twice as loud as the old model (remember, decibels are logarithmic).

The new AirTag final thoughts

When I set out to test the AirTag, I was concerned there wouldn’t be much to talk about. But after my time with it, I’ve discovered a lot to share. Namely, I’m a fan. I was already a fan of the first one, and that’s the key. Apple kept all the best aspects of the original AirTag while adding slight improvements. There was no dramatic redesign or price increase. It’s just the familiar white and chrome disc that is a little louder, higher-pitched and easier to find from quite a bit farther away.

I know some people hoped Apple would launch a credit card-style AirTag or one in different colors. It seems that Apple is more than happy to have other companies do that and participate in its Works with Apple Find My program. And I’m OK with that, too.