Hacker Newsnew | past | comments | ask | show | jobs | submit | patrickmcnamara's commentslogin

I really think this memory thing is overstated on Hacker News. This is not something that is hard to move at all. It's not a moat. I don't think most users even know memory exist outside of a single conversation.

Every single one of my non-techie friends who use ChatGPT rely heavily on memory. Whenever they try something different to it, they get very annoyed that it just doesn't "get them" or "know them".

Perhaps it'll be easy to migrate memories indeed (I mean there are already plugins that sort of claim to do it, and it doesn't seem very hard), but it certainly is a very differentiating feature at the moment.

I also use ChatGPT as my daily "chat LLM" because of memory, and, especially, because of the voice chat, which I still feel is miles better than any competition. People say Gemini voice chat is great, but I find it terrible. Maybe I'm on the wrong side of an A/B test.


This feels like an area Google would have an advantage though. Look at all of the data about you that Google has and it could mine across Wallet, Maps, Photos, Calendar, GMail, and more. Google knows my name, address, drivers license, passport, where I work, when I'm home, what I'm doing tomorrow, when I'm going on vacation and where I'm going, and whole litany of other information.

The real challenge for Google is going to be using that information in a privacy-conscious way. If this was 2006 and Google was still a darling child that could do no wrong, they'd have already integrated all of that information and tried to sell it as a "magical experience". Now all it'll take is one public slip-up and the media will pounce. I bet this is why they haven't done that integration yet.


I used to think that, too, but I don't think it's the case.

Many people slowly open up to an LLM as if they were meeting someone. Sure, they might open up faster or share some morally questionable things earlier on, but there are some things that they hide even from the LLM (like one hides thoughts from oneself, only to then open up to a friend). To know that an LLM knows everything about you will certainly alienate many people, especially because who I am today is very different from who I was five years ago, or two weeks ago when I was mad and acted irrationally.

Google has loads of information, but it knows very little of how I actually think. Of what I feel. Of the memories I cherish. It may know what I should buy, or my interests in general. It may know where I live, my age, my friends, the kind of writing I had ten years ago and have now, and many many other things which are definitely interesting and useful, but don't really amount to knowing me. When people around me say "ChatGPT knows them", this is not what they are talking about at all. (And, in part, it's also because they are making some of it up, sure)

We know a lot about famous people, historical figures. We know their biographies, their struggles, their life story. But they would surely not get the feeling that we "know them" or that we "get them", because that's something they would have to forge together with us, by priming us the right way, or by providing us with their raw, unfiltered thoughts in a dialogue. To truly know someone is to forge a bond with them — to me, no one is known alone, we are all known to each other. I don't think google (or apple, or whomever) can do that without it being born out of a two-way street (user and LLM)[1]. Especially if we then take into account the aforementioned issue that we evolve, our beliefs change, how we feel about the past changes, and others.

[1] But — and I guess sort of contradicting myself — Google could certainly try to grab all my data and forge that conversation and connection. Prompt me with questions about things, and so on. Like a therapist who has suddenly come into possession of all our diaries and whom we slowly, but surely, open up to. Google could definitely intelligently go from the information to the feeling of connection.


Maybe. I haven't really heard many of the people in my circles describing an experience like that ("opening up" to an LLM). I can't imagine *anyone* telling a general-purpose LLM about memories they cherish.

Do people want an LLM to "know them"? I literally shuddered at the thought. That sounds like a dystopian hell to me.

But I think Google has, or can infer, a lot more of that data than people realize. If you're on Android you're probably opted into Google Photos, and they can mine a ton of context about you out of there. Certainly infer information about who is important to you, even if you don't realize it yourself. And let's face it, people aren't that unique. It doesn't take much pattern matching to come up with text that looks insightful and deep, but is actually superficial. Look at cold-reading psychics for examples of how trivial it is.


On the other side of the test, I don't know a non-tech person who uses ChatGPT at all.

Another data point: my generally tech savvy teenage daughter (17) says that her friends are only aware of AI having been available for last year (3 actually), and basically only use it via Snaphhat "My AI" (which is powered by OpenAI) as a homework helper.

I get the impression that most non-techies have either never tried "AI", or regard it as Google (search) on steroids for answering questions.

Maybe more related to his (sad but true) senility rather than lack of interest, but I was a bit shocked to see the physicist Roger Penrose interviewed recently by Curt Jaimungal, and when asked if he had tried LLMs/ChatGPT assumed the conversation was about the "stupid lady" (his words) ELIZA (fake chatbot from the 60's), evidentially never having even heard of LLMs!


My mom does. She's almost 60. She asks for recipes and facts, asks about random illnesses, asks it why she's feeling sad, asks it how to talk to her friend with terminal cancer.

I didn't tell her to download the app, nor she is a tech-y person, she just did on her own.


I dislike that it has a memory.

It creeps me out when a past session poisons a current one.


Exactly. I went through a phase of playing around with ESP32s and now it tries to steer every prompt about anything technology or electronics related back to how it can be used in conjunction with a microcontroller, regardless of how little sense it makes.

I agree. For me it's annoying because everything it generates is too tailored to the first stuff I started chatting with it about. I have multiple responsibilities and I haven't been able to get it to compartmentalize. When I'm wearing my "radiology research" support hat it assumes I'm also wearing my "MRI physics" hat and to weaves everything for MRI. It's really annoying.


Thank you! I feel really dumb for not knowing about that!

Agree. Memory is absolutely a misfeature in almost every LLM use case

You can turn it off

It doesn't even change the responses a lot. I used ChatGPT for a year for a lot of personal stuff, and tried a new account with basic prompts and it was pretty much the same. Lots of glazing.

I wish AI had more access IDE-type features such as renaming or moving methods to another class. Claude often makes mistakes here because it doesn't have all the fancy tooling and indexing that IntelliJ has.

Claude recently got LSP plugins

Ah, thank you. I missed this over Christmas.

https://news.ycombinator.com/item?id=46355165



People love this graph and regularly tout it as if it explains full internet usage. Especially when they dont bother to add any explanation or comment alongside it.

This graph is mainly due to the fact that telcos use IPv6 for mobile devices, nothing more. Over time you will see that graph flatline and peter out as mobile device uage reaches critical mass.


In US even desktops have 45% adoption rate: https://radar.cloudflare.com/explorer?dataSet=http&groupBy=i...

afaik every single major US fixed line ISP is rolling out ipv6.


Yep, every ISP and every device supports IPv6 and has done for ages.

Show me one where they have disabled IPv4 and only use IPv6 that is not a mobile device or Telco ISP.


It seems more the other end of the stick: the IPv4 side of the graph is mainly held up due to corporations. The consumer internet continues to switch, but corporate VPNs are going to continue to drag down the numbers until corporations get charged enough for IPv4 address space that bottom lines start to notice.

Yes good point, I agree that IPv4 addresses are going to become a commodity in the future and their value will start to increase dramatically to the point where it is only corporations which can afford to use them. IPv6 use may well start to spike again if that happens.

Every major ISP in the US, India, and most of the rest of Asia that I’ve seen is handing out and using IPv6 now too.

Hell, chances are if you got a new router (like any new client) for your ISP, you’d be on v6 too.


Yep, and even with all those countries with their billions of mobile devices IPv6 use still hasnt even reached 50%.

Pretty much all ISPs hand out both IPv6 and IPv4 addresses to their clients, this is nothing new. When they start only issueing IPv6 IPs is when it would start truly taking off, but it will never get to that point and it will never happen.


It feels like you are constantly moving goal posts here. Your original statement was it will die a slow and quiet death. Are you now saying that this mobile use case will start to switch back to IPv4? It may not kill IPv4, like was initially planned, but it's not going away.

Apologies maybe slow death was the wrong phrase. I did mean that, but only in the non-mobile space. Obviously mobile device networks have made good use of IPv6 and will continue to.

However In another thread it was argued that when IPv4 addresses become very expensive, that could trigger a big shift to IPv6. I agree with this statement and so IMO it is possible that IPv6 may well become ubiquitous in the future.


When usage is increasing rapidly and is literally ~ 50% of the entire planet right now, how is ANY kind of 'death' a useful descriptor?

IPv4 is the one that descriptor belongs on, eh?


No I dont agree at all.

Usage is in no way 'rapidly increasing', in fact the google graph everyone is touting around shows that it has taken over 10 years to not even get to 50%. It also shows it is slowing down, the curve is starting to become less steep.

When Maximum possible IPv6 usage is not even at 50% after over a decade and the usage curve is slowing, how can you possibly say that IPv4 is dying and IPv6 usage is rapidly increasing?


Oh, now it’s a problem because it’s been about a decade?

So what, another decade and we should be mostly done?

What do you think is a reasonable amount of time to redo the entire world’s networking infrastructure across 200+ countries and 8 something billion people, exactly?

This is an absurd argument, you know that right?


> Oh, now it’s a problem because it’s been about a decade?

No, but taking over a decade to not even be half adopted does not count as rapid in my opinion.

> So what, another decade and we should be mostly done?

No, as I have said many many times, the graph is slowing.

> What do you think is a reasonable amount of time to redo the entire world’s networking infrastructure

We dont need to, thats the point. All networking equipment in the world already supports IPv6, so why isnt it at 100% usage and IPv4 is turned off already?

>This is an absurd argument, you know that right?

Who is the fool, the person saying what they think or the person continuing to participate in an argument they consider absurd?

You dont need to make everybody in the world agree with what you are saying, it is ok to have differing opinions. You know that right?

I am done now. I accept that you disagree with me and thats fine. Can you afford the same decency or will you continue to tell me I'm wrong?


Hahahahaha

According to APNIC labs, IPv6 adoption in India is ~79% and in China it is ~53%.

Those are the only two countries that could plausibly have billions of mobile devices and they appear to have reached 50%.

India: https://stats.labs.apnic.net/ipv6/CN?c=IN&x=1&v=1&p=1&r=1&w=...

China: https://stats.labs.apnic.net/ipv6/CN?c=CN&x=1&v=1&p=1&r=1&w=...


Wow, billions of devices per country and they have still only reached 50%.

Damn, the jealousy (?) is palpable. You know you will have literally zero impact on adoption no matter how snarky you are, right?

Jealousy? I think you may need to expand your vocabulary a bit!

I have an opinion on something, I assume thats ok? You seem to be here trying to prove me wrong, and also commenting on my tone of reply.

Im not trying to tell you that you are wrong, only stating what I think. If you dont like my opinion, feel free to ignore it. You are not forced to comment.


Looks like it’s right at 50% and rapidly increasing.

[https://www.google.com/intl/en/ipv6/statistics.html]

What exactly are you going on about? 5-10 years for the old devices to be EOL’d, and we’ll likely be at 95%.


Devices maybe, software won't :-\ (We're going to see ever-diminishing pockets of IPv4 around for a loooong time, much like we still see pockets of Cobol.)

pick a lane?

The trend on that graph is slowing, and when we reach criticl mass on the number of mobile devices the graph will be flat.

There is no chance we will be at 95% usage in 5 years or so.

If you like, we can make a wager?


It was simply to point out that you are objectively incorrect. No commentary was necessary. My phone and home broadband both use IPv6 primarily.

If you were correct, that graph would have been over 50% ages ago.

As it is, that graph is showing how adoption is slowing and has been for the last 10 years.

Hardly anybodies physical internet connection is using IPv6 primarily worldwide, those numbers are all mobile device space.


> Over time you will see that graph flatline and peter out as mobile device uage reaches critical mass.

...what? The majority of people access the Internet from their phone, and not only since yesterday either. Are you arguing that this is temporary fad somehow?


I don't think they are arguing for a decrease. I took flatline and peter out to mean stabilize.

That is correct, thankyou for assuming the positive instead of the negative.

I personally believe that at about 60% utilisation the line on the graph will become flat and stay that way.


I am arguing that at some point there wont be any more people without phones, meaning it has reached critical mass and so IPv6 adoption will stall. The number of smartphones in the world will not keep on going up forever.

That would only happen if all of v6's growth is coming from mobile users, no mobile networks are growing/deployed without v6, and also no users are dropping their wired connections.

You can look at the AS breakdowns on APNIC's stats and see that ASs that serve non-mobile customers are getting v6, and that some ASs for mobile users aren't. So no, it won't stall.

Slow down perhaps, but it has to slow down at some point or it'll go above 100%.


What is the source of the seasonality in that graph? Spikes up a little each summer.

Maybe iPhone release time?!

If that is the case then why does it drop in absolute terms later in the year?

Non-American countries are also importing a lot of American politics. I'd rather that didn't happen and is alone worthy of curbing in my opinion.


The new console is completely fine.


How can Hungary or Slovakia do the same? Literally, how could they? What mechanism exists?


Is this satire?


The Netherlands would be the 4th biggest state if it was part of the US.


This article isn't written for you. It's written for my mom, etc.


Surprised to see an article like that even getting shared here. The Guardian seems to be wrong on almost every tech issue.


Does your mother frequent hackernews?


This article was written for The Guardian, not Hacker News.


yet here it is, posted on hackernews


So why complain about the experts?


How is the argument much different than any current arguments? You can already get significant benefits from the state as a citizen (in European countries, at least).


The difference is magnitude, the benefits that European countries provide has made them a very attractive destination for immigration, so it stands to reason that something like UBI would make them even more so. No matter which side of that debate you're on theres a point where the math breaks down and some difficult choices need to be made, either you provide these generous benefits to your citizens or you have a generous immigration policy, but both of them together may prove unsustainable


Which is why unchecked immigration is already a problem. Increasing the benefits without dealing with unchecked immigration is going to make things worse.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: