People have been saying that for over a decade now, and their business is still going.
Your narrative depressed Google’s stock price for years, and bears finally gave up since they kept coming out with huge earnings beats.
Unless Claude adds ads, it isnt going to be sustainable for both Anthropic or you - and congrats, you've invented Google Search.
Regarding Claude, it’s true they must be losing money on free users since they promised they won’t put ads there.
Speculatively, I think people are overestimating the cost of inference for the consumer free tier chatbots. I suspect it’s effectively a marketing cost. The primary expenses are compute to train the next model, stock compensation, and heavy-duty enterprise inference (which pays for itself).
It isnt Netflix deciding, sure I'll make money, it is more shareholders are demanding profits .. and Netflx is forced to act.
So if Ads are a disease, and we (you + me + anyone with a stake in the US economy's growth) are the virus.
- fast, even the shitty AI results are nearly instant.
- some results like actors in a movie, or where to stream are useful
- instead of typing a url
But that’s it. I never use Google outside of that. It’s just ads, and most of the web is just ads or AI slop.
Funnily enough, their core revenue driver (Google Ads) is very broken too. I'm trying to run some ads, but for a week they haven't been showing due to some invisible combination of flags when the campaign was created. There's no way of knowing they're not showing from the dashboard, it only becomes apparent when you try and preview the ads with one of your search terms.
I know everyone is long Google but that experience seriously makes me question how valuable their ad business will stay in the future.
1) training data (common crawl as one example web data source), and
2) live data optionally retrieved at runtime.
My comment was about 2), and that part runs via a search engine (Bing?), at least if you look at how ChatGPT does it.
So, why would you use Google as a tool or search target when you can, in some combination, go direct to the website (or whatever the target data endpoint is) yourself as an LLM provider to retrieve the most recent data or rely on your own "hot cache" of that data that was crawled recently but said data is not stale enough warranting a live web crawl to retrieve and present to the user or AI agent? Is this capability to perform retrieval from a data source in real time not similar to an AI agent?
Broadly speaking, I'm just spitballing on the concept of "You must use a search engine for an LLM to return 'live-ish' results" as I think we're directionally headed to where that isn't the case.
I'm not sure frontier labs do it, but imo fetching live data requires fetching both the search engine (1) and the website/page (2). The search engine gives you search results + content snippets (potentially stale), the website/page gives you the actual content fresh from the source.
The tradeoff I see here is between liveness/staleness and cost (hitting an index if of course cheaper and faster than querying live websites again).
And that's where you're dead wrong. Absolutely wrong.
I would actually recommend using a LLM over normal search for that sort of thing, because SEO has ruined those high intent phrases. The AI equivalent of SEO is also a problem there, but it's so much less prevalent.
I personally like google over asking an AI. Esp. since something like google is a substrate without which an AI would be far more prone to hallucinations.
Some of their newer efforts like Waymo and AI hosting for Apple could turn into large businesses to make up for any lost search revenue.
They also have significant costs to maintain their search revenue, such as $20B a year to Apple alone to make Google the default search engine in Safari.
If search becomes less lucrative, they would presumably pay less for such deals.
So yeah, definitely a time of disruption for Google and they need to keep moving fast on many fronts, but they are still well positioned in multiple markets to be successful.
Curious to see how long that lasts - YouTube and Google Maps were fairly mainstay apps for iPhones for a while, and eventually Apple cut the Maps cord to do their own thing. I don’t know if that’s from an existential concern of “we need to eventually move Maps in-house”, but given the Google of it all I wouldn’t be surprised if Apple is already at least planning on how to do the same with AI once the bulk of the usage evens out i.e. let someone else worry about it now while also getting a read on what rolling their own would feasibly require.
But if they are successful in making on-device models that work well, then you don’t really need as much cloud infrastructure.
A model serves a different purpose from a web crawler.
I'd prefer to be able to find source material over steering a model I have no control over while managing hallucinated outputs without the ability to fact check.
Just because it sourced some materials using RAG doesn't make its outputs valid, accurate, or factual.
Is it though? As far as I can tell they continue to maintain total domination of web search, and LLMs do not replace search engines, they work on top of them.
So is IBM.
There was a pretty interesting article in the Wall Street Journal a few weeks back about how IBM is flying under the radar, and doing well by not taking the hype bait.
It noted that 70% of all credit card transactions on the planet go through an IBM system.
The problem is that those COBOL systems have to be absolutely provably correct, for both financial and regulatory reasons. LLMs can't do that. They're designed to be variable.
You can't vibe code a bank transaction system. "Close enough" isn't good enough in some fields. A minor glitch in a video game may result in screen artifacts. A minor glitch in a banking system can crash the economy.
COBOL on mainframes has worked reliably for decades. On the scale of what a bank will spend on operations the mainframe doesn’t really cost much.
Banks are sort of the canonical example of “no one ever got fired for buying ibm”.
I'm not sure what they do these days - it's about 20 years since I used an IBM Thinkpad to connect to an IBM pSeries.