Exactly. Which do people think users will be more accepting of: “we sold all your data and chats for 300 million” or “we got paid 300 million to integrate AI”.
This is a data sale, pure and simple. xAI certainly isn’t making their money off their LLM/inference. OpenRouter shows Grok at about a billion tokens a day while the top 20 account for 2.5 trillion per day.
I’ve suspected Elon expanded his failing strategy to include data brokering when he saw the opportunity to get access to everyone’s data via Doge. Hence the reason Elon is ready to step away now that Doge (many of which are xAI employees) has finished gaining access to the data from every government system. Quietly offering access to corporate and political clients to query the data of every single person via Grok seems like an easy way to generate some revenue when no one wants your AI.
I’m sorry, I don’t think this is on Musk or xAI. IMO it is on Telegram. Only Telegram has the obligation to know what their users want and look out for their own users. It’s not unreasonable for Grok to want to have more users.
Who knows how much they’ll make off the data they obtain from users and chat history data, but in this day and age, no one gets paid to integrate AI, they pay for it which means there is clearly some other goal. And I can tell you there is a reason he didn’t make this offer to Signal instead, where they wouldn’t have access to any chat history, and that it would mean xAI (at this point which looks more like a data harvesting company) becoming a “partner” would normally allow data to be handed over. In this case, however, that isn’t necessary since lTelegram’s privacy policy is basically, “we can use your data for whatever we want”, including selling it for 300 million and claiming it is the payment to integrate AI.
Are you claiming it's a stupid business decision, or something else? Because they must be thinking that there is a profit to be made somewhere, either through training data or brand exposure to Telegram's 1 billion users.
Or Musk just want to say “hey, we growed in users by x%” this year so that he can keep pretending being in the race, for some street cred', like the way he pays some progamer in Asia to boost his PoE2 account.
That's quite a viable possibility, I don't get why you're getting downed.
Grok is indeed trailing the race and could use extra numbers. It doesn't matter to their bottom line (or their ad campaigns) if these numbers come from a flat rate agreement with Telegram.
If anything, Telegram should have bargained for more.
Yeah, I'm not contesting that at all. I think Telegram sold its soul a long time ago, so I kind of just treat it as a self-contained socially oriented splinter-net with its own integrated app environment... and it's actually really cool and feature rich. I also find Grok to be an LLM that performs pretty well, so I don't mind the integration, much the same way I didn't mind the AI integrations into WhatsApp, but I understand that many people just want a messenger that's a messenger so I can understand the pushback.
I thought that. Then, I remembered PayPal straight up paying people to use their platform. We saw how that worked out.
So, the question is whether they should pay to generate demand in a new market. Then, who to pay and how much?
I'll also note that OpenAI took the market by offering an expensive service for free (ChatGPT). Then, they offered a monthly plan that may or may not have been profitable. One could argue that OpenAI has been paying people to use its service for a long time.
I also wonder if xAI gets something out of it. For instance, they might get all the conversations with the AI. I believe ChatGPT similarly put free, user conversations to use internally in ways that boosted their paid apps. xAI might have some plays like that.
When was PayPal doing that? I remember having to create an account 20 years ago because they were the only payment processor option on Ebay, that seemed to be their moat. These days, I use Paypal as a "wallet" that's connected to my CC, to pay for stuff on sites where I don't want to put in my CC information.
> If this is something their users actually wanted
Users wouldn't specifically ask for Grok, but they might like to have access to an AI assistant. When you are competing against OpenAI etc. it makes sense to incentivise big, PR generating customers such as Telegram.
That shows that their trust in users selecting google as primary if given the option when setting up a new device is not high. Otherwise they’d negotiate it down.
I think the point is that users mostly don't bother selecting their search provider, many will simply use what is pre-configured. Google is paying Apple so nobody else will.
I don't think users really want this for what it's worth, but that's a bad argument. The median (social media) internet user pays exactly zero for anything, so deriving expectations from money is a bit silly. The actual currency to watch is attention. If they don't like it, engagement will drop.
That’s the industry consensus. Sam Altman talked about something similar on a Stratechery podcast.
It not exactly unrealistic to imagine that people will ask LLMs about flights, hotels, restaurants or even insurance.
All that needs to happen is for LLMs to add a “buy now” button, and for the provider to take a 15% commission (still a lower take rate than Expedia FWIW).
I could use some help understanding this better. If you ask an LLM about flights or I think pretty much any of the categories of services you cite, what exactly would I expect to get back?
The "dream" is that the LLM/AI does all the work for me and just magically gets me to "this is the perfect flight for you and here are the reasons" but I have to tell it the things I care about (price, time, etc.). A lot of the time it's not exactly clear, even to me, which of those parameters matter the most so as a consumer it's actually nice to see, for instance, the options with times/costs/stops/etc. and for me to be able to look those over and make a decision. The LLM could provide those options, but then what has it done that I wasn't able to do with Google Flights or another ota? Is it just using more natural language in the request? Or do most people really want to wholesale handoff the decision and just go with the "trustworthy" LLM/assistant without any of the rationale?
I suppose actual personal assistants do that type of thing all the time for wealthy people, but that doesn't seem like it would be applicable to the masses who want/need to comparison shop for the best deal that meets a bunch of criteria.
I say this as someone who gets major value out of LLMs already, but for buying things in particular I'm struggling to understand why you'd want to hand off the "browsing" and just fast forward to the "buying."
I see there as being two main types of people who’d benefit from using LLMs for booking flights.
The first are power users, in the sense of people who have complex requests. Like those who’d ask for the cheapest flights from Manchester to Turkey during the summer school holidays for without layovers unless those layovers were in Paris, for a holiday to be roughly 10 days long, excluding very early morning flights. Such a request could be made with existing OTAs, but would be painfully time consuming.
The other type of user for whom LLMs might be useful are the opposite, those with very loose requirements. Think “get me a flight to a warm place tomorrow”.
Skyscanner etc would still have value for the people in the middle.
> When you use grounding with Google Search, and you receive Search suggestions in your response, you must display the Search suggestions in production and in your applications. For more information on grounding with Google Search, see Grounding with Google Search documentation for Google AI Studio or Vertex AI. The UI code (HTML) is returned in the Gemini response as renderedContent, and you will need to show the HTML in your app, in accordance with the policy.
Even if all this true, there’s unlikely to be a mass switch campaign until an alternative with better privacy and security gains the smoothness, flexibility, and platform coverage that Telegram has. Most people I know who use it do so because it runs everywhere, is nice to use, and isn’t a Facebook product (which while perhaps not being as bad as a foreign government, is a more prominently perceived threat for most people).
I really don't get the sentiment of everything Russia. Russia has the GDP of a province of China. Russia does not have a thriving manufacturing sector. Russia has increasingly less influence or power over the world. Why would powerful people in the US succumb to to Russia's influence? That just does not make sense. If anything, maybe the Lenin-loving Stalin-admiring Marxists would love Russia, like half of the university professors in the 50s loving the Soviet Union, to the point that they'd rather leak all kinds of information, including how to build atom bombs, to the Soviet Union. Oh, and New York Times even won a Pulitzer for praising the Soviet Union by the great Walter Duranty in the 1930s.
The "why" is definitely an important question, but it's certainly pretty obvious that views of Russia are relatively favorable among some on the right [1].
> Russia has increasingly less influence or power over the world.
Russia punches above their weight because people believe what they want to believe (and what their peer group they want to fit in to believes), and are easily corruptible by "small" amounts of money. And they have nukes.
They have nuclear bombs. A lot of them. And the person who controls them has little to lose in using them. So we have to be nice and keep that person happy.
That's diplomacy? That does not mean that high officials and the richest people in the world bought Russia's propaganda machine, if any, or worse, get bought by Russia, right?
Well the side effect of everyone playing nice with Putin is that he's most likely a multi-billionaire, possibly even in the top 10 richest in the world.
That actually looks like forcing their view, censoring you just because they might not like your position. Useful, you cannot do a thing. Your comment is simply gone. I might wanted to save it, but it’s no longer.
Yet, they did not bother to remove my useless neighbouring comment, only flagged it. I’m not surprised anymore. Not the first time I see this behaviour.
That's just not true. If Telegram wanted it a small amount, and xAI wanted it a large amount, it would be normal for Telegram to negotiate to get paid for doing it.
This isn't quite the framing I'd use—the key thing that would make it make sense for the deal to go this way is if there's competition bidding for the slot. Say OpenAI also wants in and was willing to pay for it. Absent competition for the slot and assuming that Telegram actually wanted the deal, we'd expect to see at most a no-cost exchange.
There's no reason for Grok to pay this much for the deal unless either Telegram sees it as a net negative that needs reimbursement or they're competing with other bids.
What if Telegram sees it as a $100m benefit but they know that it’s A $500m benefit to Grok. If you were in telegram’s shoes in that situation, would you just charge nothing, knowing that Grok would happily pay $300m?
It’s his play to be an even more important part of the military industrial complex and the surveillance state.
He already has rockets, internet satellites, social platforms, he has the ears of the president and now he will probably have a backdoor to one of the most popular “encrypted” chat apps.
> Telegram is the most popular messaging app in Iran and Uzbekistan [1]
> Telegram’s largest market is India, which accounts for more than 20% of its userbase. Telegram also has a large amount of users in countries with heavy censorship and surveillance, such as Iran, Russia, and Uzbekistan. [1]
> Percentage of users via region: Asia 38%, Europe 27%, Latin America 21%, MEMA 8% [1]
It sure is valuable data - if you are a three letter agency. Not sure if it's valuable business data.
You don't know what you are talking about. Read on the multi-year attempts of Putin trying to ban Telegram and Telegram fighting back (technologically and also politically - demonstrations, etc). Telegram won the battle and Putin had no choice but to admit defeat.
Similar things happened in other non-free coutries as well.
Many Telegram groups are only Play Store banned or banned access is restricted based on the user's phone number. This is why you must install the APK from their website directly instead of using an App Store version.
This have nothing to do with Apple. One of largest anti-war public channels in Russia by relatives of mobilized people was marked as "FAKE" for a very long time by Durov's personal decision.
Durov serves kremlin as much as any other company that operates in Russia.
It is ridiculous that Apple decides what Telegram users can or cannot see, especially banning harmless things like nudity but allowing cruel things like war videos or propaganda.
WhatsApp isn't E2EE by default either, since default flow pushes you to backup your key to Google Drive.
Signal isn't E2EE, given the security blunder in which private images from your gallery were sent to random contacts (which indicates a scary state management situation in the apps, this isn't easy to do). E2EE implies that you purposely send content to specific people which is encrypted, not that your app sends potentially embarrassing or intimate pictures to your boss behind your back. That blunder is unforgivable.
>Telegram won the battle and Putin had no choice but to admit defeat.
That just seems such a unreal claim. Telegram removed features like "people nearby" after its CEO was arrested in France. Who seriously believe that the kind of threats France establishment would employ on such a person could dwarf those of Russia establishment in term of bending the braves?
You should not forget that before creating Telegram Durov was the head of VK and he left it to the government with all data, photos (including deleted ones) and messages.
I'll warn that the FBI was publicly trying to get warrants for information while they and NSA were siphoning it off in secret from the same companies. One was likely a cover for the other.
Unless there's legal protections, assume in your threat model any company has let their host government, maybe others, backdoor their offerings. It might have been willingly or forced. Police states like U.S. and Russia should be assumed to subvert any pprovider.
If they don't like that, they need to repeal the Patriot Act, ban requiring companies to attach black boxes to their internal systems, give companies immunity for publicly talking about court orders, require companies to disclose what data they give to the government, and let individuals know what was ordered after a period of time. Then, I might trust statements about what they do or don't share.
Also, if these bother you, try not to commit crimes.
> I'll warn that the FBI was publicly trying to get warrants for information while they and NSA were siphoning it off in secret from the same companies. One was likely a cover for the other
That’s a wildly ignorant take, but it makes for a nice conspiracy theory if you make no effort at all to understand the legalities.
Different warrants authorise the collection and use of information for different purposes. FAA 702 warrants only authorise targeting non U.S. persons outside the United States for foreign intelligence purposes, where there is probable cause to believe the U.S. person is a foreign power or is an officer, employee, or agent of a foreign power.
The FBI has criminal investigation and counterterrorism functions which relate to persons in the U.S. and/or where there is no connection to a foreign power. They obviously need different warrants to authorize those activities.
That's what they said before the Snowden leaks. The Snowden leaks and latter revelations showed they were lying.
For example, they use a different meaning for the Word "collect." Instead of interception and storage of data, collect means an analyst looked at it. So, they technically weren't collecting U.S. citizens' data if analysts hadn't looked at that specific data yet. Technically... based on a strange definition of collect.
They originally also said this was limited to terrorism. Later, data showed they were looking at many more crimes. They were also passing the data onto many agencies. They were told to use "parallel construction" to deceive people about how they got that data.
Finally, BULLRUN and ECI-classified level showed they were weakening U.S. security standards, but pretending to strengthen them, so they could attack U.S. systems in secret at any time. Per "Core Secrets," they were also having U.S. companies give them backdoor the FBI could "compel" them to make (somehow).
With all that, they were caught lying under oath repeatedly. They got criminal immunity for that, too. I don't believe one word they say at this point. I also assume they're doing the same things they repeatedly lied about before and for which they can't be prosecuted.
It sort of used to be, although not exactly as a messenger - it was never good for one-to-one private conversations, but as a social network with channels and groups.
However, for the last couple years enshittification is in progress - it's not at Microsoft Teams levels yet, but they're really trying to get there, shoving more and more ads (third- and first-party both) into users' faces with increasing frequency.
Since the protocol is actually open, you don't have to use the official client.
I'm curious as to why you say it was "never good for one-to-one private conversations". I've been using it for this exact thing for many years, and still find it the best option currently on the market for a variety of reasons (e.g. unlike Signal it doesn't limit the number of devices which are linked to the same account).
I wasn't a telegram user for long so this may have changed but isn't the secret chat limited to device to device? So messages sent from my phone can't be decrypted by the same user account on my laptop?
I just remember thinking, well this is dumb, and going back to Signal (Signal annoys me in other ways, requiring a single phone to be the "master" and other devices to be merely linked. I miss keybase, they had a great system including paper backups)
The main Telegram enshittification is caused by spammers and scammers. They purchase lots of Telegram Premium credits and proceed to spam the hell out of users and they get away with it. Telegram Premium users are treated much more leniently than regular users when it comes to moderation affairs.
> True. Agreed in principle, but formalities are pending.
This is how I know I'm not cut out for business - because I just can't lie with total impunity. Like, how could you do a global, public announcement like this when there is no dry ink, and the other side obviously hasn't authorized you to announce, and feel literally zero shame ("formalities are pending", lol).
I feel like more and more one of the biggest requirements for success in business (not to mention politics) is just total shamelessness.
He is the CEO of a corporation. Let's assume Pavel and Elon had a phone call and agreed on the deal. One of them or both are entitled to announce it. Pavel Durov announced multiple deals before on his tg channel. Now it is surprising turn of events, it seems they didn't work out details to the degree implied, but calling it lying is a far stretch.
> Let's assume Pavel and Elon had a phone call and agreed on the deal. One of them or both are entitled to announce it.
Lol, because it was Elon himself who called out Durov's original tweet with his "no deal has been signed" response. So obviously the agreement was not agreed upon yet.
Ironically this sounds just like Elon's infamous "funding secured" tweet. I'm tired of all these bullshitting bullshitters, and then the follow on of apologists.
Obviously one of them though they had an agreement and other didn't.
Or maybe Elon is backpedaling after being hand-slapped by US of A administration for dealing with what they potentially consider to be unfriendly social media network.
<< I feel like more and more one of the biggest requirements for success in business (not to mention politics) is just total shamelessness.
I am going through 'Signal and Noise' by Nate Silver and, after dismissing it at first read, now I am starting to wonder if his attempt to categorize some people on how well they can predict things is onto something. The boorish bravado ( or at least a perception of it ) is clearly part of the requirement though. You simply cannot even register to people with all the noise around them.
Call me stupid, but I fundamentally do not grasp why xAI would pay Telegram for access to its users. $300M seems like a crazy amount of money to put into a chat app.
I am genuinely curious about the business strategy behind the move bc that would be a market worth exploring - having something that the AI industry would pay for bc they are willing to spend a lot right now.
When OpenAI bought Windsurf a lot of the discussion was about which market and which customers they hope to get access to by that acquisition. This deal gives xAI privileged distribution among nearly a billion of users (and potentially future customers) for a magnitude less of money.
Who will „win“ the LLM-AI race is as much undecided as is the common way to interact with them and this seems like quite a sensible bet on distribution for a huge userbase with a very specific integration into a platform. Doesn’t seem at all crazy to me.
$300M is peanuts when it comes to pumping fake valuations of AI companies. Elmo will now boast about 1.4 billion active xai users. He might even claim those are paid users, because he paid them to use his ai :D
AI training data is in short supply from new sources. Chat and group apps are handy for many reasons, but one obvious one is people posting images with caption text they provide basically does your image tagging for free (remember the original big image model was just a list of URLS with useful alt-text descriptions).
On the chats tab, right above the green "+" floating button there is a "meta AI" icon on my version of WhatsApp. If you click on it it opens up a chat with the AI like normal.
I have a premium account as I really like the idea of just paying for the service I use.
I highly assume that the AI will be integrated in a way that you can mostly disable it or barely notice it if you aren't looking for it, like all the other paid or special features.
So many people never noticed stars, NFTS, a whole nft market, pay for messages, pay for groups, ... While using it ever day
I had my doubts before, but this was the proverbial drop in the bucket: I removed my account. People can still contact me via phone, sms, email, or Signal, or by just ringing at the door.
So that means Telegram users will use XAI for free, which means it will cost xAI potentially multiple times of $300M a year to serve the Telegram users?
In retrospect, I agree. I worded it too strongly out of emotion. "Everyone could be the product of a service, free or not" would be closer to what I think. The message being that if being a product is important to a would-be user, they should investigate in every case, because power dynamics play more of a role, than being free or not.
Telegram channels are upstream of ALL text based social media and to an increasing degress, culturally upstream to many content trends as well.
Take the war in Ukraines for example, the uncensored and real time updates you get in open Telegram channels make most intelligence agencies except for Five Eyes nations look regular.
This deal may be much bigger than it seems off the bat. The cohort of people using Telegram to exchange content is maybe the top 5% of the world in many important niches.
the good news is that when this doesn't yield any positive results for xAI and they need a bailout then they'll get bought by Tesla, and Elon is saved once again.
Effectively, of course, Twitter and Solarcity went bankrupt. That's what really happened.
I hope the CxO's at companies realize this, and so realize that Musk's big plan for Twitter ("just fire everyone, keep collecting the income") had a slightly different outcome when put into practice: fire everyone, rehire half, get publicly shamed by important people refusing to go back, bankruptcy in 2 years, 5 months. In other words, since 2022 Musk lost 5 million per day, on average, for 2.5 years.
This is a self-comforting argument I keep seeing all over against Tesla. Yet, their refreshed models are showing up on roads very quickly.
Demand for Tesla isn't permanently going anywhere just to make a few new Elon haters happy. I also don't like Musk, but I am not deluding myself like so many these days that Tesla is dead. Elon is much like a roach; he will survive nuclear Armageddon.
> not deluding myself like so many these days that Tesla is dead
Nobody said it's dead. Just that the car part of the company, the stable part, is somewhere between being abused and neglected by Musk. While the self-driving part has no competitive advantage and the robotics part seems to have legitimate synergies with xAI.
I can't help but think this means that all my telegram conversations will now be fed into grok. I'm curious how I'd go about verifying that this is or isn't going to be the case.
I'm thinking about most of my communications on Telegram and it's kind of hilarious to imagine it suddenly starting to reply to people on Xitter with unsolicited horny furry roleplay instead of unsolicited white supremacist rants.
Not so hilarious that it doesn't make me want to consider trying to convince all my circles to move to a community-run Matrix server or something though.
Yeah, sounds like they finally accepted that tweets don't make LLM smarter, only less biased, and so they're doubling down to feed more tweet-like data to make it smart and biased.
Did you have an expectation of privacy using Telegram? I think that's the real issue here. If you were paying for Telegram, then I would say yes, those conversation should not be handed over to xAI. But, if you were ostensibly not paying for it, then I think it should be assumed that use of the service is considered consent to take that and use it as they please.
So yeah, it's almost certain that everything you ever put into Telegram is in Grok.
SMS is never claiming to be E2E nor is any army of SMS defenders online talking about how some virtue of SMS is almost same as E2E. While I don't like Telegram I will happily admit it is better than SMS, but how is that an argument for anything.
this is actually pretty bad deal for russia.
with xAI.
US based company will get access to the messaging of all russians/ukrainians/Iranians (with soldiers being very well represented), with the ability to run NLP and identify information of great importance to the NSA/DoD/CIA.
Imagine if AT&T routed all Americans' SMS/text messages to the Russian version of xAI, would you call this a bad deal for Russia or USA ????
The $300M "in cash and xAI equity" could turn out to not be that much money, depending on what percentage of that was equity.
xAI recently raised at an $80B valuation and it's highly debatable if it's actually worth that...and who knows if you'd ever actually have a liquidation event.
Fair enough. I'm not at all familiar with SpaceX's comp but I assume they do private liquidity events/stock sales for their employees. Who knows if X/xAI will do something similar, and if outside equity holders are subject to those same rules. Far above my pay grade to figure that out.
If paying Apple 20 billion to be the default search engine in Safari, paying Telegram 300 million in cash and equity for access to their 1 billion monthly active users, seems like a good deal for xAI. Where else are they gonna get users at that scale?
xAI is paying to be "the default" AI of s chat app of 1B monthly active users.
They pay for distribution and they likely believe they can later monetise this by sheer mass, roping people in Grok or offering professional paid services.
Telegram has in-app monetisation for 50 payment providers and more than hundred countries. They also have crypto but it is unrelated.
Also if someone still believes into the freedom of speech, Telegram was just banned in Vietnam this week when they refused bend the knee to the government. It aligns with Musk's ideology.
This just made a whole ton of people I know on telegram (furries who make the internets go) immediately accelerate their movement off the platform. I found out about this news because of some backup signal rooms started sending notifications. They need to get ahead of the privacy implications and quickly. Furries are basically the roma of the internet at this point - what they value hasn't changed (anonymity, moderation tooling, rich expression, security) but they move on as platforms betray their original usecases. They're like the canaries for enshittification especially around privacy.
Too early to tell - usually they'll settle on something with a balance of those main values. If matrix is truly the new IRC then perhaps that, though it doesn't seem mature enough yet and it seems like it might be costly if one needs to host non-text content. Furries contain both highly technical people and also very low-tech artists so any platform that requires a lot of maintenance can be a non-starter. Signal seems like privacy over expression. Maybe this is the time that they'll make their own.
Most of that crowd that I interact with was already off Telegram to begin with, and (in some cases, begrudgingly) run their own Matrix instances.
I think if you interact with non-tech-users, Signal is the go-to; Matrix has way too many thorns for normal users to put up with in comparison to the alternatives.
As a furry: lmao no Matrix. Matrix does not get the job done at all. Profiles stored on the server you decided to join? Lots of silly design decisions.
Also the whole "furries make the internets go" is not really that true. There are certainly many of us in tech fields but going to meets/cons the average furry is not working in technology. The whole "suspiciously wealthy furry" thing just arose from the divide between tech worker and retail worker within the community. Gotta remember that most furries are like 16-25. Many tend to drop out or leave the community over 30-40 or so.
Not a furry but before anything happens I guess we need to wait how telegram is actually going to implement it. The way it got presented, basically as additional add-on not getting in your way for normal telegram business at all I doubt it will matter.
I don't understand how Telegram is still around. The app has zero support, even for paid users. You can't rely on it, and unlike Meta products, where at least there is someone to escalate to, in Telegram it's nonexistent. My account was taken over by a malicious actor, and I can't get it restored no matter how many emails and DMs on X I've sent them. Zero response
As someone that has friends on every chat platform out there, I must say that Telegram is the most polished chat app in existence. Everything just works. I notice bugs in WhatsApp all the time, but hardly ever with Tg. Even stuff like video chat and video messages has always been reliable and fast. As a developer, the API is amazing... in 30min you can make a bot and publish it. The real shame is two fold: chats are not encrypted by default and there's so. much. bot spam on the platform. These two problems have made it my least favorite platform, even though the usage experience is the best.
And it has been the most polished chat platform since a decade, they have been just excellent at creating a product people love to use. I really don’t want xAI, so I guess I could switch to WhatsApp, but that would be a downgrade…
Because it's the only social network that mixes blogging/feeds (channels) with anonymous chats and public groups. Most other messengers require your phone number to be exposed (although WhatsApp has been fixing it lately).
Is this the same Grok that started inserting talking points about "white genocide in South Africa" on completely unrelated queries? Even if you wanted some AI chatbot integrated with your chat app (which I personally don't), it feels like Grok is the worst one you could pick.
That is absurdly low given that grok probably will be learning from the inputs. If anything this number should be in the billions. Do telegram users even want this? I thought it was privacy focused app.
could be used for spying purposes, if xAI can analyze telegram chats and identify texts of interest to the next-get USA military industrial complex (SpaceX, Palantir, Anduril, etc)
It's the next logical step after companies shoving AI into every corner of their own products regardless of whether their users want it - now they're paying other companies to shove AI things into their products regardless of whether their users want it. Genuine user interest doesn't come close to justifying their insane valuations so they have to put their thumb on the scale by shoving it everyones face and then pretending that's the same thing.
See also: Googles AI summaries, which always get top billing so they can tally nearly every search up as an "AI engagement" regardless of user intent, and can't be disabled because that would get in the way of what's clearly the actual goal (to juice the AI metrics as hard as possible, user experience be damned).
It's absolutely wild and scary watching how much money is being spent on pushing AI down the unwilling public's throats. Nobody wants this. Yet we're hiring expensive AI researchers and developers, buying datacenters full of GPUs, and now paying "partner" companies, to deliver this thing that nobody is asking for. What in the world is going on here? What am I not understanding?
> What in the world is going on here? What am I not understanding?
I don't think you're wrong in any way. I've been in denial for the past few years because the world is going crazy with AI and politics. But it's actually very good for me because I'm shunning all that shit and I focus on local people and local problems more: taking care of the finances of a non-profit, being more available for my friends and relatives, solving actual problems that people may have, etc. Denial is great and it makes more active. The downside is that I now have the calendar of a CEO and less time for me, but I believe the world need some care and we all can do something about it by doing small stuff.
What's not to understand? Enormous amounts of money have accrued to a tiny proportion of humanity in the past 30 or so years. There is no way there wouldn't be tons of waste when spending decisions are made by so few people.
Now add in the fact that these decision makers are often openly avaricious egomaniacs who don't even make symbolic efforts help the poor and vulnerable, that narrows the scope of their spending to wasteful, sometimes outright harmful investments.
> Enormous amounts of money have accrued to a tiny proportion of humanity in the past 30 or so years. There is no way there wouldn't be tons of waste when spending decisions are made by so few people.
the other issue with that isn't just the decision making but the fact so much capital is accruing at the top they have nowhere else to put it all, meanwhile average people are struggling to pay rent and buy food...
Reaganism set the wheels in motion but those wheels didn't actually come off until events like the dotcom boom normalized billion dollar valuations for half baked MVPs, creating a generation of future nutters like Thiel, Bezos, Zuck and Musk. Things accelerated even further with zero interest rate policy post-2008, making capital free for this "job creator" class while working people were charged "market rates" for home and education loans.
But the land they're grabbing is desert with no water and no access roads. Does anyone besides the few with their wealth invested in AI believe that AI is the next iPhone moment?
> Does anyone besides the few with their wealth invested in AI believe that AI is the next iPhone moment?
It doesn't matter because for such an industry-wide hype, there are no consequences for being wrong. If a CEO ignores AI and it does become the next iPhone moment, they'll be deposed in short order. If "everyone" is wrong and nothing comes off AI, they'll write off some investments, write some "What we learned" LinkedIn posts, and carry on. Our existing framework has no incentives to correct or innoculate agaisnt hypes led by the management/capital classes
I think it’s probably as simple as some old fool on Sand Hill Rd got suckered into writing a check for this nonsense with promises of world domination by AI’s promised infinite profit with minimal cost. And to keep the whole charade going, everyone has to pretend that this will eventually see some returns otherwise the whole farcical system will come crashing down. We can only hope that happens and some correction rears its head.
The end result being all of us suffers in some way for the greed of a handful.
The corporate world is overrun with executives designing products that look like solutions to other executives but that don't solve any problems problems people in the real world actually have.
It is funny seing xAI, the trash-tier AI company, integrate with Telegram, the trash-tier messaging service.
Enshitification is often a company-wide culture problem, but the fish does rot from the head.
There are a variety of reasons why a company might begin to over-incentivize short-term gain (or high-stakes risk-taking) at the expense of customer happiness and possibly to the detriment of the company's long-term interests.
For example: Growth stagnation, an existential threat, a pessimistic long-term financial outlook, bad reward structure, low customer regard, organizational infighting, low employee retention, etc.
The sudden emergence of AI and volatile economy are triggering several of those for a boat load of companies. And, well, show me the incentive and I'll show you the outcome.
Example: Since this deal was cash + equity I wondered whether telegram has a public valuation. I searched google and got an AI summary saying that the market capitalization of TELEGRAM is $7.4767.
That's dollars not billion dollars, because google's AI summary was referring to some scam coin which has a total marketcap of a big mac and fries plus or minus. It seems now to have updated to refer to the messaging app and _their_ (probably also scam)coin.
xAI has essentially zero market outside of twitter, and with recent system prompt shenanigans from Elon, I cannot imagine anyone signing up to pay for API hits when there's a non-zero chance your application will suddenly start complaining about white genocide. They're painted themselves into a corner with the product and Elon's increasingly erratic behavior, they now have to pay companies to use their service.
Are they paying for API hits and developing application stacks around the service? It's clear who uses it, it's a lot less clear who might be willing to pay for it.
Because it isn't disabling them - that would be an option to actually disable them.
Of course there are hacks round it, but ultimately Google want us to have the AI summaries, don't care about user choice, and the best course of action is to change search engine.
or instead of searching for "best dog breeds for apartments" change it to "best f'in dog breeds for tiny s-hole apartments" - feels much more cathartic
We've seen multiple companies in the last twelve months blast past any past benchmark of fastest growing company ever. It's become pedestrian for some of these companies to scale to $10m ARR in a quarter which has never happened before.
"Genuine user interest doesn't come close to justifying their insane valuations" - classic HN copium
> The ChatGPT integration, powered by GPT‑4o, will come to iOS, iPadOS, and macOS later this year. Users can access it for free without creating an account, and ChatGPT subscribers can connect their accounts and access paid features right from these experiences.
Models are only as good as the data that is fed into them. OpenAI is paying Reddit 70M for access to the data. So the real value here is the conversations, not the model.
And in the case of Telegram, you will get very intimate data about people. You know in real time who they are talking to, what they are talking about, etc... Its extremely valuable data
I think this is more the case. xAI is looking for more data to ingest.
I'm not sure how the integration will work with Telegram if the contents are supposed to be "secure". Are you just allowing your conversation to get exfiltrated to xAI? Does the other party you are talking to get a say in that?
I don't like Grok that much, but there's nothing particularly new or interesting about this deal. Telegram has a big international audience, it makes sense someone would pay to be the default for user adoption reasons.
"Feels oh so telling that Google has to pay Apple and not the other way around. If Google search was so great wouldn't we all be clamoring for it to be the default search in Safari?"
The Google Search deal seems like a much more defensible business decision to me for a couple reasons. For starters Google gets revenue almost every single time that search integration is used so there is a direct return on the investment. What's the conversion rate for a paid Grok account? It might turn out to be terribly low and $300M is a substantial amount of money. I doubt a normal company without Elon's vast wealth network that needs to actually make money would gamble on a deal like this.
The other reason is that the Apple deal is a big part of maintaining Google's search monopoly. Owning a tech monopoly is vastly more valuable than competing in a crowded market so locking down market share to achieve that can be worth spending more per user than their median value. Grok isn't even in the ballpark of an LLM monopoly so those benefits don't apply in their case.
No matter how good your offering is, you always need marketing if you are competing against an established behemoth in the space like OpenAI. Even Google has been unable to make much of a dent in OpenAI's daily active users despite having superior distribution and (recently) comparable quality.
Grok 3 is legitimately one of the best general purpose models out there. People don't know about it because ChatGPT is "good enough". And people have no reason to care unless Grok 3 is 5x better or ships a feature that goes viral like Ghibli portraits.
Telegram is ostensibly a competitor if you buy into Musk's pitch for X as an "everything app". Paying Telegram $300M instead of developing chat features and/or marketing proves that plans for the "everything app" are dead in the water, or perhaps were never sincere from the get go.
It's honestly a good fit. Telegram also doesn't have the best image, but people don't care, so it's highly likely that they also don't care about the image of XAI or Musk.
I do struggle with the 1 billion users, but I also don't believe that X has 600 million users actual users.
We'll have to see how privacy is handled (whether you can opt out of it suckling on every one of your personal conversations) but assuming some baseline decency from Telegram, I don't think anyone would be "forced" to use it.
> We’re side stepping the elephant in the room: X/xAI/Musk’s brand is toxic and forcing the product onto users is one of few paths available
This statement is baking in a lot of personal convictions, even if they feel self-evident. Telegram has a billion users and not everyone one of them will share those views. This setup is a lot closer to Google paying Apple and Mozilla to be the default search engine than some desperate attempt to get people to play with your toys.
I know my experience is anecdotal, however, I know at least three people at my workplace who have said in our Slack channel they will never, ever but a Tesla because of Musk's antics while typing it using a Statlink connection.
It sounds like they'll buy an alternative product when one is available. Tesla, X, and xAI all have multiple viable alternatives - superior alternatives even. That's not yet the case for Starlink in many locations[0], but in 5-10 years, your colleague may be using an Amazon Kuiper, One World, Garmin[1], or Apple[1] connection.
0. Symmetric gigabit fiber Internet remains the gold standard, where one can get it, but unfortunately that's not many places.
1. I hope these companies are at least looking into doing these, as it's adjacent to their current products.
Grok3 could be leagues ahead of anything else in the game and yet a large portion of the western population would never use it due to its owners antics.
True, but Musk is an exception. And deliberately so, his media image was one of the things that made Tesla so notable. But we’re seeing recently that it goes both ways.
Musk was the PR, press and Marketing departments rolled into one for Tesla and X, so much so that he fired the PR team at Twitter.
"The public doesn't care who the CEO is" is certifiably not true for Musk ventures, and Musk exploited that brand value in the past positively. The inauguration Seig Heil, and the subsequent DOGE misadventures are the other side of that "personal brand" coin, which very much exists for this CEO.
Don't underestimate how far they've fallen. Tesla is nowhere near the top in Europe anymore. VW sold three times as many EVs as Tesla, with BYD in second place.
This is wrong and I would encourage you to check your sources or look into the industry more before posting a blind yahoo article that was clearly written by AI.
In many of the comparisons, they try to make it more dramatic than it really is by saying volkswagen as an entire brand sold X more than tesla. Well of course they did, they sell hybrids and ICE vehicles too.
> Tesla is nowhere near the top in Europe anymore.
This is factually incorrect. They are #2 overall BEV sales in europe for Q1 2025 and also own #1 and #2 spots for best selling BEV models.
I would expect such a low quality comment from reddit, not HN, but alas here we are in 2025.
And additional reminder for folks that have an axe to grind that perhaps clouds their judgement of reality: when you've dominated the BEV market for almost 10 years, going sideways and down in marketshare is pretty much the only option.
Definitely makes sense that if they are a market leader for an extended period of time they would expect to trade sideways; my question is whether or not Tesla is expected to come out stronger as a growth company in the upcoming quarters (i.e. their recent talks of doing robotaxi work; are investors seeing that as a serious growth angle?) or should Tesla begin investigating an income/dividend stock angle?
Valid. And I do think they'll continue to own the lion's share of total sales for a while. But is there any concern over their being down YOY for Q1 while other manufacturers are largely growing? Or is the expectation that it's a temporary blip and they'll continue growing in the upcoming quarters from previous YOY numbers?
> And I do think they'll continue to own the lion's share of total sales for a while
Maybe in the US where they are protected from Chinese competition. Tesla's fall in Europe has been precipitous, falling by more than 50% YoY. BYD sales overtook Tesla for the first time ever in the last quarter in Western Europe.
Don't bet against Google. Their cloud is growing despite coming from 4th or 5th place, now it's on 3rd place and growing faster than AWS and if I recall correctly, faster than even Azure.
> "Incidentally, have you got a moment to talk about the white genocide in south africa?"
Is this some sort of joke, or are you genuinely wanting to discuss what's happening here in South Africa?
If it's a joke, I think it's in very poor taste at the expense of a persecuted minority. I have seen it, experienced it, and know full well how real it is, no matter how much social-media pushes it one way or the other.
Stages currently active in SA: 1, 2, 3, 4, 6. Some people claim that the high murder rate among white SA'ns means they're also in stage 8/9, but actually the murder rate among black SA'ns is even higher.
There's no genocide, but there's a whole lot of red flags and generally terrible things. Just because Trump believes something does not make it wrong.
From elsewhere on that site: "Dr. Gregory Stanton, Founding President of
Genocide Watch, warned that early warnings of genocide are still deep in South African society, though genocide has not begun."
There are countless examples of early warnings. Here are a couple:
Many similar statements have been made by other politicians, among them Julius Malema, who collectively represent a large part of SA's population, over the course of many years.
Don't respond to something like this with a flippant statement about clutching your pearls.
You need to make your way through 1-5 to get to 6, but I specifically excluded 5 (organization of militias).
There are rumours of this but if it's true, these militias must be very secret indeed.
One problem is that violent criminals run rampant and the state seemingly has little desire to stop them, and zero capacity to do so. Meanwhile prominent politicians sing songs whose lyrics include machine gun sounds and calls for the murder of whites.
Parallel to this, there are job restrictions limiting the maximum number of whites companies may employ or promote. Franchises limit the number of white owners. White business owners are strong-armed through law and government contracts to give up some of their equity. There's regular talk about seizing white-owned property.
Whatever label you want to put on all that, I think the it's fucked up.
The worst part is, the average poor black South African is innocent in all this and now has to live in a place with a spiraling economy, power & water outages, even worse crime than the whites have to face, terrible standards of education, and much more.
>You need to make your way through 1-5 to get to 6
This isn't true. As the genocide watch page says "The process is not linear."
White South Africans are less likely to be victims of crime, make more money, live longer, and are over-represented among corporate and political leadership. Taking steps to undo the damage done by apartheid isn't fucked up, it's necessary as evidenced by the aforementioned inequalities.
And I'm telling you that 1, 2, 3, 4, 6 are the ones that are active right now.
> "Taking steps to undo the damage ... it's necessary"
What a bullshit argument.
How does any of that help "undo" the damage done by apartheid? The government of SA prizes ideology over outcome. The ideology is to attack Western things, even if the outcome is that black South Africans are now very much worse off than they would've been if SA was thriving.
And to cite this as a defence for the abuses and outrages that has literally landed white SA'ns - despite their relative wealth - on a genocide watch.
>And I'm telling you that 1, 2, 3, 4, 6 are the ones that are active right now.
I trust genocide watch more than you to determine which of the genocide watch stages of genocide are underway,
>How does any of that help "undo" the damage done by apartheid?
White South Africans make up 8% of the population but own 72% of farmland. This is a direct result of apartheid and colonial racism more broadly. Expropriating some of this land and returning it to black South Africans is directly undoing this damage.
Thank you. This is all that needs to be said on this subject.
The word genocide does not mean "a lot of red flags and generally terrible things", nor was it made to describe that.
>Don't respond to something like this with a flippant statement about clutching your pearls.
Don't clutch pearls then.
As long as you keep abusing the word "genocide" to apply it to the plight of white people in South Africa, you'll get the response you think is flippant, and I consider to be insufficiently stern given the harm and disrespect of such usage.
Saying this as a Jew whose family members were killed by the Nazis during WW2, by the way.
They were KIA as soldiers, so I'm hesitant to label them as victims of genocide, even though they were certainly the target of it — out of respect both to them, and those who didn't get a chance to die fighting.
You don't get to call a demand for respect flippant.
> Thank you. This is all that needs to be said on this subject.
No? There are tons of completely insane things happening in SA, and much that needs to be said and done. If you're saying the minimum threshold for caring about what happens in other countries is actual genocide, then I disagree with you.
People have been incorrectly calling this a genocide for a long time, which is why the PDF from Genocide Watch dates back to 2015.
On the right you have people trying to make this even worse than it is, and on the left you have people trying to ignore it, or minimize it.
>There are tons of completely insane things happening in SA, and much that needs to be said and done. If you're saying the minimum threshold for caring about what happens in other countries is actual genocide, then I disagree with you.
I'm not saying that.
The subject at hand is whether there's a "white genocide" taking place, and that subject is summed up in a single word: no.
>People have been incorrectly calling this a genocide for a long time
Which is why it's important to not perputate this harmful falsehood any further.
>on the left you have people trying to ignore it, or minimize it.
By using the term genocide where it's not applicable, you're actively minimizing the actual genocides that have taken place (or are taking place) — and by extension, you're complicit in minimizing the very issue you're discussing.
See, we both agree that whatever is taking place in South Africa is not as bad as an actual genocide.
But by using the word "genocide" in conjunction with it, you're diluting the meaning of the word reserved for the absolute extreme — you're helping spread the notion that genocide doesn't have to be that bad; that "red flags and terrible things" fits under the something sort of kind of like genocide label.
What we have in the end is the parable of the boy who cried genocide [1].
The point of the parable isn't that there's no threat of a wolf attack, nor that is shouldn't be seriously considered.
Main difference is Google was paying a rev share, so both parties make bank off the deal, mostly Google. Here, there is no shared interest, just one company taking money from another for distribution with no financial return. Neither company has any incentive to make it work well.
Google didn't have to pay anyone back in 1999 or early 2000s.
Google paying Mozilla and Apple to be the "standard" search engine absolutely means it is a bad product. IMO it's purely anticompetitive too, but I'm a competitive market radical.
I feel similarly for how chrome didn't win by being "good", it won by being bundled, and by putting one click "Hey hey click here" buttons on google.com.
Google pays a rev share on the return it gets from the deal. This isn't pay and cross your fingers child's play here. Google makes most of the money for search transactions that click ads and the distributor, Apple, makes a small fraction of that for providing the search access points. Where's xAI's lucrative return here? Where's their 80:20 split?
Time for some no-evidence conspiracy theory: the API-based "AI"s are surveillance vectors, so this is just cover for running all your conversations through a tool to determine which ones to report to which authorities.
(while in theory they can be run locally, in practice this is rare)
Most users would probably use Google by default if forced to pick on install. Google is effectively paying so that their competitors like Microsoft won't sign their own deal and become the default Firefox search instead.
That's different as X is buying a form of exclusivity there.
Paying a browser to become the default search engine is just to serve first. And, there is the argument users are not asking for, nor want Ai features, I bet most don't want it. They just want to chat, with humans that is.
They are inserting Ai everywhere they can, not to rank up their product as the kind of product people are looking for, against those of the competitors.
A browser without a default search engine is a downgrade for everyone. Although it would be more ethical if it simply prompted for which to default to. One could point at it and see an issue, but that's pretty different.
Yeah but chat results in paid subscriptions and it disincentives telegram from creating their own AI chat. Google also pays for most browsers and other ecosystems which aren't really considered "locked"
No different. It is an attempt to prime the pump to stoke valuations. The product is the stock price go up, the illusion of value creation is the work. "Please Use."
If people aren't paying for it, and you have to pay them to use it, what is the value? Russ Hanneman Silicon Valley pre revenue rant meme here
Google pays Mozilla a fraction of the money Goggle makes from ad clicks driven by Firefox, a revenue sharing agreement. Where's xAI's 80:20 split here? Where's the default AI interface in Telegram that users would fill with some other AI service if Telegram wasn't getting paid for the distribution? Where's the user expectation of AI in the product in the first place.
This is nothing like the Google Mozilla (or Googlge Apple or Google whoever) deal.
"Feels oh so telling that Google has to pay Mozilla and not the other way around. If these offerings were so great wouldn't we all be clamoring for it?"
Please tell us what you believe this is telling us. You’ve left it up to interpretation and this feels more like a Reddit quality comment than one that the hackernews community deserves.
In my opinion, it’s telling us that competition in the LLM space is accelerating and it’s anybody’s game right now. It’s important enough of a space that companies are willing to pay for exposure and squeeze out the competition.
I’ve interpreted your comment as “llm has no value so the providers are paying for people to use it”, which is, naive at best
> “llm has no value so the providers are paying for people to use it”,
I feel like for that to have been even possible to understand from that comment, then "not the other way around" couldn't have been there. If authors opinion was that LLMs has no value, why'd Telegram pay to use Grok?
Maybe I'm just used to reading between the lines, but I think parents comment strike a fine balance between saying too much, saying enough and saying too little. It's understandable what they mean, if you read the full comment. Not everything has to be explicitly spelled out for the lowest common denominator.
They pretty much did say “not the other way around” with
“If these offerings were so great wouldn't we all be clamoring for it?”
The entire question is flawed. It’s a rhetorical question and the implication is “these offerings are not so great and we aren’t clamoring for it”.
I’d argue we are clamoring for it, and we have a lot of options here and they are all great options.
But since the rhetorical question is flawed you now have nothing to anchor on to know exactly what OP meant because if they are flawed here that means whatever they meant in the first half of the comment is also likely flawed.
Maybe you’re not as good at reading between the lines, or the line in general, as much as you think you are and you’re actually the LCD that would be served by a deeper comment.
The whole point of the comment is that it is “telling” of something but the commenter at this point could come back and basically go in any direction. That’s an indication the comment is not saying enough…
That's an interesting way of looking at it thanks for sharing. In that way it's similar to Google paying to be the default search engine in Firefox or on iOS. They _did_ have the best search engine at the time but they still paid platforms to have it embedded as a default option.
if people don't want to use the AI it means the AI is not useful or interesting enough. paying 300M won't really move the needle in that. Now , if they actually paid the users to use it ... who knows
Contrary to popular belief that's not really so much a thing. Even back in the day RP was only vaguely popular with a portion of the community. Most furries don't have sex in fursuit, only some.
Otherwise it's not yiffing, it's just sex. Probably more prevalent among furries given that we have a higher ratio of queer people than wider society, and we're all mostly more open about sex. Straight non-furs are comparatively Catholic in their relations.
Sad to say this as a formerly happy user of telegram:
Boy that tracks. The shadiest AI model provider paying the shadiest chat app provider to integrate the racist slop machine into the scam and CSAM distribution machine.
xAI engineers, I genuinely hope you get the bag soon and switch jobs back into a reputable AI lab.
Out of interest, why were you using telegram to begin with, considering that it's known for being a haven of shady content? Not just hate groups and paramilitaries, but also tons of pirated content? All that content browsed gets linked back to your phone number does it not?
It was a reasonable messaging app for a lot of folks in ~2013-2016. The stickers support was excellent and it had a lot of excellent features, even though its psuedo-open source stance was always a little weird. I also ascribed some level of trust to Pavel Durov as ultimately being anti-authoritarian given his history with Putin and thought that was ... fine, I guess. Anyways, with the rise of crypto came a lot of other unsavory things where Telegram became the de facto comms platform for criminals. While criminals will use any medium, it's also a matter of critical mass to me.
Not OP, but Telegram is the only popular option with E2EE that has good clients across all desktop and mobile OSes (including Linux), doesn't tie your account to your phone number, and doesn't limit the number of devices which can use the same account.
As I said before here recently; I am replacing all my tooling by stuff I built and that includes chat. I own my company so my colleagues can leave or use it too (they help me build so yes, they will use) and family will come on board if they have a side access to people who won't, which we do: we integrate other chats and for those we losen the rules. I have no interest, when I die in 40 years, to depend on even one line of software not inspected by me.
Telegram is a haven for scammers and malware authors, who frequently use it as a command and control channel. These scammers and malware authors can now seamlessly integrate Grok into their tools.
What is the alternative? Discord is just a PITA to use. I need to suggest something to a group of people that have been meeting online since 2007 (it's a small group with widely varying technical abilities/time).
It is worth noting that WhatsApp Trust and Safety team is more effective in removing and blocking large-scale scam operations compared to Telegram.
For instance, Telegram’s founder was recently arrested in France for failing to adequately remove malware, scam, and CSAM from the platform. It was only after his arrest that Telegram began to take moderation seriously, although their efforts remain woefully inadequate.
The fun fact is that the supposedly uncensored Grok ("unhinged") is actually very much censored. The only thing it can do is to say bad words and be anti-Woke, but it actually refuses to be really subversive or give its opinion on many topics.
The Instagram pages the article references almost all link to...you guessed it...Telegram, where they can distribute illegal materials with less oversight.
How so, when they could have paid around $300M for the data?
Telegram may even be more valuable in terms of conversations, because Twitter data has very low signal:noise ratio - such as replies whose entirity are 3-word sentences or just emojis
I don't know about you but many of my messages are 1-3 emojis. Not that it matters, that is an easy thing to filter out of training data.
Elon's election-buying effort was certainly assisted by the Twitter purchase. And look how few (formerly) blue checks have left the platform despite his antics.
Definitely a good investment in taking over the presidency of the United States, but I don't feel like many bluechecks are still on Twitter? The platform feels dead, you can't even access it without logging in.
