Google is probably screwed without it. Even if ad-tech isn't exactly a bubble, I strongly believe it is overvalued even still.
Furthermore, the product that made their name is thoroughly lost. I dread Google Search today. If I found something on Google yesterday, I have no idea if I'll be able to locate it through the sea of blog SPAM and SEO garbage tomorrow. Google web search strongly favors big websites when sometimes the page you need is never going to be there. As an example, Google puts heavy weight on near-SPAM sites like Pinterest which produce almost no original content, and on abusive hosts like Fandom (formerly Wikia) who manages to beat out community wikis by replacing admins that try to migrate off and winning in SEO.
I am not trying to suggest Duck Duck Go or Bing are actually better overall, but if I see a web search covered in results like that, I sometimes find that they do in fact give me somewhat better result rankings.
The only thing I really have any strong affection for is YouTube, and the people and content that made the platform special are constantly unhappy with it's corporate direction, not to mention the utter lack of support, claiming and affirming in appeals stupid pointless crap like "you said a naughty word in the first ten seconds of the video" when anyone with working ears can easily verify they did not.
And that's the tip of the iceberg. I think a lot of ventures, like GMail, only ever made sense as computers and bandwidth just kept getting dramatically cheaper. It's difficult to beat free. Well, what happens when those things stop getting cheaper, and GMail keeps getting more expensive to run?
> The only thing I really have any strong affection for is YouTube, and the people and content that made the platform special are constantly unhappy with it's corporate direction, not to mention the utter lack of support, claiming and affirming in appeals stupid pointless crap like "you said a naughty word in the first ten seconds of the video" when anyone with working ears can easily verify they did not.
This is mostly an aside, but it's worth putting into perspective that many of these complaints are related to Google's ads and automatic monetization through ads, as well as their content recommendation network/userbase.
In other words, I think as Google's ad-revenue shrinks (or the bubble bursts), the Youtube ad policies will seem to become more and more draconian, but despite all the anger towards Youtube, the alternatives involve... no (automatic) monetization at all.
I don't think a lot of people are really conscious of how a lot of modern Youtube (specifically the big content creators who have made careers out of Google's monetization options) are implicitly reliant on Google's ad-tech bubble. And most people are pretty derisive of Google's attempts at alternative monetization e.g. youtube premium. I fear when the bubble bursts (or just shrinks a lot), a lot of the high quality content people enjoy on youtube will burst or shrink with it.
Big YouTubers have seen the writing on the wall, and have been diversifying income streams for years. A lot of artists are migrating towards the "patron" model (think Renaissance patronage, but crowd-sourced) using tools like Patreon or Liberapay. This strikes me as inherently more pure than throwing videos at Google's algorithm to see which one can bring home the biggest slice of automatic-ad-revenue pie.
The attention economy is doomed, too many people are online now for it to be viable. Patronage and connecting with followers will come to dominate. Not to oversimplify, but there's a theory that all an artist needs to be financially stable is 1,000 true fans chipping in a few dollars a month. Google's YouTube middleware currently occupies a comfortable niche in the bringup process required to meet that figure... but they're ultimately nonessential.
As the bubble shrinks, creators will look back and wonder why they ever bothered driving millions of random people toward a platform that mostly pays out to Google.
I agree with most of what you say here, but a few small points to contest:
First, it's not just big YouTubers moving towards the patron model. I see many smaller creators doing the same, and I'm some cases thriving.
Second, I don't agree that the patron model will become the dominant model. I think this line of thinking is more of a medieval model from which the word derives. In terms of the number of people paying by ads & PII or through patronage, the ads& PII crowd will win. (the dollar value per contribution might be lower).
Third, I don't have any reason to believe that creators will "look back and wonder" why they would like to attract a million viewers. I think it might be a very good way to advertise to their patr(e)on-age.
Finally, there is another revenue model of direct sponsorship / placed ads which you do not seem to consider. I think it is a major contributor to the economics of medium-large creators.
Maybe my gut feeling is a response to your notion of the "attention economy being doomed" because "there are too many people online". To me, it seems the opposite. The attention economy is thriving because there are so many people online.
(Interesting topic! Thanks for making me type all of this)
Yeah, the 1000 fans thing can totally work, depending on what your market looks like. $10/mo * 1000 is $120k/yr, which is a perfectly reasonable stable base from which to pursue a creative career (music, art, indie game dev, writing).
And if you can make it to 1000, it's pretty easy to make it to 2000, and then more - of course just getting to that 1000 can be a long slow challenge to be sure.
> A lot of artists are migrating towards the "patron" model
If ad revenue shrinks, making ads less useful for them (or even if not, given the patronage trend) why wouldn’t Google just enable the patronage model as a monetization model within YouTube, unencumbered by concerns for ad-based (or general audience) monetization.
Going to a three-tier (1) free/nonmonetized, (2) premium (with or without ads as an alternative to viewer-pays-google flat rate), (3) paid-subscriber-only model for content (where any creator could particpate in any combination of tiwrs with different content), where the last would be free of most of the content monetization, or even just content, restrictions that apply at the other tiers, would seem to be a fairly simple and natural evolution of the service. And you could easily have the Google share at the last (and even the middle tier) ratchet down with longevity and established user base so you'd pay a bigger share when you benefit most from the platform for discovery, and less when you need it less.
Google has already created channel memberships, where the channel can offer perks like member-only content in exchange for additional payment to the channel owner. It certainly looks like an attempt to squeeze out Patreon and similar services.
The catch is that Google gets a 30% cut, compared to 3-5% for Patreon and such.
Good point, I forgot to account for the platform fee, which they charge in addition to the payment processing fees. So a creator on the Pro plan pays 8% just to use the services plus processing fees (2.9% plus $0.30 if over $3). So it seems like a typical creator would pay around 12%.
> Google's YouTube middleware currently occupies a comfortable niche in the bringup process required to meet that figure... but they're ultimately nonessential.
I disagree with this, though I'd love (and genuinely hope) to be proven wrong. It's hard for me to imagine starting from nothing and building up a profitable following the way people have on Youtube without Youtube's userbase and infamous "algorithm".
It might just be a matter of semantics, though. I agree that content/media existed before Youtube and will exist long after it -- but I'm not sure Youtube's niche of non-traditional content creators without the money or social networks traditionally needed to publish media could do so in its absence.
More prosaically, there's a big gulf between starting out, and reaching 1,000 true fans throwing some money at you every month. Youtube plays a pretty big role IMO in bridging that gap, and without it, only people with traditional support networks or in higher economic classes will be able to afford "doing content creation" full time.
I feel like a common response (and one sort of echoed in another reply) is "Good, let's bring the old internet back, when people did things for passion and not money". I sympathize a lot with this, but I also think a lot of content I enjoy is the product of people putting tons of time and effort into making "good content" -- time spent doing pretty miserable and uninteresting work. So I'm inclined to think the end of the "youtube era" will bring the end of a certain class of creator.
Not that I think this is preventable, since like others here, I think the whole ad thing is unsustainable, and we're seeing more and more writing on the wall. Hopefully Patreon + co. and more decentralized content aggregators like Mastodon can fill the void well enough.
> the alternatives involve... no (automatic) monetization at all.
This is 100% what it boils down to. Google is obviously not the only advertiser in town, but if the ad bubble pops, so does the free content economy on the internet. Some people will argue the internet would be better because of it, but the transition away from ad-supported viewership would decimate the web we know.
I'd argue the internet I liked most was decimated by the current free content economy, though.
I'm not saying I disagree with you by any means; a lot of what we would lose would be a great shame. Only the strongest creators could probably find a way to stay alive, likely through more direct sponsorship (both Patreon and well, traditional direct sponsors. You already see this a ton on YouTube anyways.)
I get that the internet will never be as scrappy and experimental as it once was, but the way monetization spread rewarded a lot of people who thrived off of essentially taking free stuff and monetizing it. I mean a lot. Think of "prominent" "content creators" and "influencers" like Fuck Jerry. In the process of doing this, I think all of the fun was milked out of everything. Today's internet feels like a giant strip mall.
I really do wonder if this is the best middleground we can get.
Of course, if ad-tech does turn out to be more or less a giant bubble and that bubble pops, we lose everything anyways, without any chance to transition, and all of the anguish was for naught. It feels like quite a gamble to keep hinging on this so strongly for the best content online, so I hope that the patronage model continues to grow.
Google is the only advertiser in town that matters. There's some private clubs you can go to (Facebook's properties, Amazon) but for generally placing ads or selling ad space on the internet it's Google.
I use search for coding a lot, and lately I've noticed it's hard to find what I'm after. If I'm looking for some small thing like what an sftp command looks like, I don't want an SEO essay about why SSH is great and we should all use encryption. I just want the example line.
This seems to be what ChatGPT can give me. It doesn't always work, but it also doesn't beat around the bush.
100%. I have found it sooo difficult to trace framework/library documentation down that 3-4 years ago was first few results. Something has ruined their algorithm.
Yup, exactly the same experience here. It feels like there's been a recent wave of SEO-optimized websites that consistently pop up when I google specific errors, but rarely have literally anything of substance.
That is odd, what exactly were you looking for? In my case Google results are still top notch. Sometimes I search for SO question every X years and I’m always able to find it easily.
> I have no idea if I'll be able to locate it through the sea of blog SPAM and SEO garbage tomorrow
If I have to sift through one more useless AI-generated/copy-paste article where they spend half the damn thing explaining the most basic stuff and then still not answering my question (or giving a simple "yes/no"), I'm going to lose it. It's not even just technical answers. I have to basically type "...reddit" or some forum I like to get ANYTHING useful anymore. It's even just basic media questions.
Example: "Is there an [IP sequel] being made?"
First 20 results: "Is there an [IP sequel] being made? Read more to find out. First what is IP? IP was released in 2012. It was made by such and such starring such and such as they go on an adventure to do the thing..." 4 paragraphs later "There is no information on a sequel. Thanks for reading!"
Yes I've learned to spot these as have most folks around here I imagine, but holy hell they will eat 2-3 pages of results sometimes! It's infuriating!
This kind of problem isn't just because of search, it's also because of content creators. I believe what all content creators quickly learn is, on any given topic, the funnel is extremely wide on the most introductory knowledge and then becomes uselessly narrow as you drill into specific info.
To use a specific example, the other day I was trying to learn more about programming shaders. There is an unending number of basic level tutorials: what is a shader? There are probably 500,000 people searching for that topic. But if I want to figure out a UV scrolling texture, I'm screwed, there's almost nothing. Because the audience for that is maybe 500 people.
There's just no financial (or even reputational!) incentive for anyone to write an article / make a video on something requiring deeper expertise. There is stuff out there, of course, created by the tiny minority of passionate enthusiasts or dedicated professionals. Google is really searching for a needle in a very, very large haystack. And I think Google (and other search) has been more and more challenged by that over the years as the internet moved from a niche enthusiast place to a mass market place. Particularly when all the hay is spending massive amounts of effort on trying to pretend to be a needle.
tl;dr I despise using Google (or DuckDuckGo) these days but I don't think all the blame should be assigned to the engines, they're actually facing a tough problem. (Not that you personally were saying that, but this is mainly a thread of hating on Google.)
>This kind of problem isn't just because of search, it's also because of content creators.
See I think this becomes chicken/egg real quick too, though I do think Google is more to blame. Content creators are responding to the incentives google created. Think back to when sites would have all sorts of keywords buried in metadata and/or have blogs that were complete nonsense full of keywords that weren't easily findable on their sites. They did it because they were rewarded. So what did google do? They penalized it, and voila! It stopped. This was like...2014? Maybe earlier?
Ultimately if google wants better search results they need to go back to the drawing board. They need to disincentivize blog spam.
Yeah, but penalizing people based on metadata or keyword lists is incredibly easy. Compare that to trying to figure out which recipe site is keyword stuffing vs which happens to have a story by someone who really did journey to Tuscany to learn the perfect cacciucco recipe.
Sometimes I've fantasized about Google having human editorial that would evaluate these content farms as they grow and make the nuanced decisions of which to derank. But then, if they really did that, they'd get sued or regulated pretty quickly.
I feel weird arguing for Google's side since I actually do have a lot of criticisms and generally dislike Google search. I just think this content farm issue looks very thorny; humans are built to deceive each other and at some point of sophistication that gets very difficult to control.
Hell just threaten it and make a few examples of folks to dissuade it. A business isn’t going to risk it if they’re worried they could be the unlucky ones. The ones hitting the top search results aren’t randoms making hundreds of sites, it’s often just C-tier blog sites that depend on dishing this stuff out regularly to get ad traffic.
I’m not saying I have all the answers here but I feel like this shouldn’t be too hard to at least make a dent in.
So, I googled Kagi, found their link… Not only is it required to create an account to use their service, they want a credit card to bill as well! Hard pass.
Yes, as a paid service who directly sells their primary product rather than a data mining, anti-privacy homunculus whose most famous product is merely a means to an end, an account and payment is required. This is strictly better than Google.
Their founder is also on record as "down ranking misinformation", which is not a judgment call I want my search engine making. Relevance and truth are two different things.
If you are into self-hosting, I hear Searx is pretty good. Been thinking on setting that up on my personal server so I can just stop relying on companies
The problem is DuckDuckGo is very USA and English focused. I've switched to it myself long ago, but it's very frustrating when I search for things and it brings up American shops higher than anything else. I'm not going to 5,000km to go shopping. You move out of English it's even worse. For general tech searching it's on par with Google. But if you are looking for things locally I often find the need to switch back to Google.
That's not how businesses process payment. Any responsible finance department will call a hard no on that. (And from a consumer side, asking to pay with a gift card is usually a huge red flag that a scam is incoming)
Yep. Let those below argue over “they want a credit card!” as if that isn’t exactly what people have been begging for. I gladly pay for Kagi and people who are now complaining because it doesn’t have some esoteric payment method were never looking to have the problem solved in the first place.
I’m not sure what these people are searching for but I want to pay for a search engine that doesn’t show ads and doesn’t track me. A credit card comes with the territory for mass adoption.
They said something like this after I exhausted my daily search allowance during the beta: go search somewhere else. I mean who are these disrespectful little pricks/snotty kids? :DDD
Interesting - haven't heard of Kagi before but I looked them up and I can definitely see how a subscription-funded search engine leads to a better incentive structure. But can you provide qualitative examples of how your experience has been better than on Google?
- Result relevance seems on par with Google about five years ago. Very rare that I need to go past page 1.
- Advanced search operators actually work. When you quote something it actually searches for that exact string, when you minus something it actually excludes it.
- Categorization in the search results; for instance all of the low signal listicles can be segregated into their own section.
- Image search that is not intentionally nerfed that will just give you the image directly
- The ability to up and down rank specific sites from your results (something you can't feasibly do without an account anyways); good riddance Pinterest and spammy stack* reposters
- DDG-like bang searches, with the ability to add custom sites
- The obvious one: no ads or tracking or perverse incentives
Have been using Kagi since the invite phase.
The moment they opened up and switched to subscription, I immediately subscribed (you'll run out of the free quota in ~3-4h - 50 searches per month).
Paying 10$/mo for something you do hundreds of times a day, and getting consistently better results than Google, Bing or DDG, doesn't seem too much to me. I happily gave up D+ subscription to have Kagi instead.
It's very rare for me to have to scroll past for the first 5 results with Kagi. With the other search engines, nowadays I can't find anything relevant in the first page.
Im thinking on going with Searx instead. I don't want to pay money, and I already have a server I use for a variety of reasons. I get why most wouldn't want to do that...self hosting isn't for everybody. But I would prefer it to paying someone else or having my data sold
Honestly, It's the same price as copilot, which is imho one of my biggest tools in my coding arsenal currently.
I don't think it should be priced as high. I think they should add other services maybe w/ their own prices, like a photonmail competitor, and api's so people can build things off their tech but pay per query, etc.
I think they could at least have tiers like 100/month 1, 400 searches for $2.99, 1000 for $5.99, unlimited for 8.99. They'd reach a bigger audience, which would likely mean more dollars overall. If I were them, I'd also create a mobile app that's completely free, just to get people hooked on the results while using mobile, or maybe have some privacy-based ads in the app, though the actual search results would have none, and remove that for paid accounts.
> Furthermore, the product that made their name is thoroughly lost. I dread Google Search today. If I found something on Google yesterday, I have no idea if I'll be able to locate it through the sea of blog SPAM and SEO garbage tomorrow.
You're ignoring mobile + local + timely searches which are incredibly common and very valuable. Google doesn't put "heavy weight" on Pinterest or "SPAM and SEO garbage" there. These SERPs are full of local and useful results and paid ads/placements. This is where Google's interests lie.
By the way sites like Pinterest, dominate Google results, a stranger would incorrectly believe that it is some kind of key "go to" website of the internet.
Instead it's the reason why ordinary people need to learn negated search terms.
-site:pinterest.com -walmart -amazon -alibaba
Devil's advocate, lots of Google services are 100% benevolent. Like Google Open Fonts. Did you know most of the font licenses packaged with Windows, you're not even allowed to package them into a game?
The article doesn't seem to back up its headline at all.
Google is under antitrust investigation for ads, while it talks about the threat to the company being AI. The main point of the article is:
> It would be a lot easier Google could move fast and break things... But with the full force of the US government bearing down on it, the search giant can’t do that now... Google will need to conduct itself slowly and tentatively in order to avoid jeopardizing itself in its battle with the Justice Department.
Sure it might need to move more slowly in its ads business (not a good time to buy an ads-related startup). But I see no evidence Google will need to act "slowly and tentatively" in terms of incorporating chatbot functionality into Search.
So I don't see anything "screwing" Google about this at all, not in the least.
One of the potential issues a lot of people are speculating about is the lower impression counts. For example - when trying to write a script using some api's I'm not super familiar with, I might do 6+ google searches. This will be similar for a lot of fields when trying to find an answer to something, news, other research, etc.
That means you're loading a fresh page of ads every search AND all the ads on those sites too.
With chatGPT (And I'm aware Google has something similar) - I can (and have) just asked it a question, and it will answer/write entire scripts for me. That means there's only a single page load.
Basically people are speculating Google may struggle to monetize a chatbot to the extent they can monetize every single search + site load. You certainly _can_ the speculation is whether the drastically lower impression counts will impact their overall profitability.
Honestly I don't think it's going to make much of a difference.
The real money is in ads that are targeted to a search -- whether you're asking Search or a chatbot for "best fleece half-zip for men" you'll get shown the same ads targeted to that.
While if you're asking it to write a script for you... you probably were to busy to be even looking at or clicking the ads Google was showing anyways back when you were looking for Stack Overflow answers, no matter if it took 100 searches. (There's also nothing stopping Google from showing you 3 new ads before each new chatbot response.)
I have full faith in Google figuring out how to insert just as many ads as before. ;)
> I have full faith in Google figuring out how to insert just as many ads as before. ;)
Oh they will I'm sure - I'm just not sure if they'll make _as much_ or not. Basically the question is if they can recover the cost per impression money to cost per click IMO.
Or maybe they pivot to something way worse - like in-content ads. Oh man imagine "best fleece half-zip for men" - "The most highly recommended and best option is the [insert marketing here but make it sound great and natural]".
But yeah I highly doubt their existence is in jeopardy or anything like that.
I'm not so sure AI is as big a threat as Google perceives it to be; back in 2009 the HUGE EXISTENTIAL THREAT was Social, therefore Google+, but that came and went and Google's growth continued unabated. That said, I do see a place for language-model driven search. And I shudder to think that language models will be specifically coerced, bribed, and have ingrained in their brains only what advertisers want. If you thought the search landscape of the 2020s is fully of SEO'd crap, just wait. Google is not in an enviable position.
I agree about AI, though for slightly different reasons. Google published the foundational paper for some of the most important tech behind the exciting AI stuff[1]. Because there's every reason to believe they are technically capable of competing, the reason they aren't offering AI products is probably based on their beliefs about how practical that is right now (cost / reliability / etc). They also have all their Waymo experience.
Basically, I'm not convinced that Google is "behind" on the tech, but just disagrees about product strategy. Obviously there are lots of examples where this dooms the current leader (Xerox Parc!) - but I think it's too early to say? I wouldn't short them yet imo.
If Google is so afraid of AI, then they should have done the most Googley of things in building AI into Google to get everyone use to it, then drop it. Deprecate the crap out of it like all of the other things. Then, since Googs dropped, it must not be good right? Boom! problem solved
TBF Google found they already owned a first-tier social network. It was disguised as a video sharing site.
The article quotes someone as saying history repeats itself. Debatable. Google built Alphabet to be defensible against antitrust cases like this. Expect a consent decree.
As you wrote: "I do see a place for language-model driven search." And unlike social media this has less dependency on user perception of whether a social network is cool.
I honestly don't think that chatgpt will destroy Google either...whenever I use search engines, I use them to find a website I could not remember the name of, locations like restaurants, etc. ChatGPT will be a tool to be used alongside search engines, not a replacement
It didn't turn out to be existential, but they absolutely lost the Social "war" and it's costing them billions: Facebook owns a large chunk of all ad revenue now, in no small part because of this.
Not that I'm complaining - the less that's under a mega-conglomerate, the better things tend to be. But it was and is a gigantic failure for Google, possibly their largest ever.
>Facebook owns a large chunk of all ad revenue now
Facebook rode the wave of open 3P consumer data which is now closing everywhere, from DMA to iOS restrictions. That's why they're trying desperately to pivot to the metaverse, so they can use first party data. Facebook is currently experiencing an existential threat.
Social media diminished to algortihmic media (e.g. TikTok) which will likely diminish to AI media sooner rather than later.
Definitely - fast ecosystem shifts are the nature of social sites and I'm glad the market is re-learning this... to some degree anyway.
Google can still point to it and say "billions missed out on" though, and not a single person in the board room will disagree.
Meta-wise, I'm glad the acronym is now MANGA, and I have my popcorn ready. The metaverse is the second biggest joke I've seen this much money spent on, second only to NFTs. I quite like VR, but "owning" "the metaverse" is unbelievable nonsense.
But Alphabet will probably be fine. YouTube and Android are still billion-user products that aren't being displaced with AI, and their advertising business doesn't seem threatened much by this either.
So what's the scoop? I use ChatGPT, it hasn't replaced my search engine. They're insisting that search is obsoleted with AI (and Google is spooked internally) but I have yet to see any research-based evidence supporting it. Personally, I use both tools for completely different things, and while they are similar by-way-of a text box with a prompt, their responses are completely distinct and neither one can replace the other.
I recently asked gpt to give me a markdown table of the continents, their landmass and their populations. I checked the numbers and they matched wikipedia.
To be fair to ChatGPT though, all the search engines first recognize that the query is trying to perform an arithmetic calculation, then pass it on to a calculator. ChatGPT already understands what is being asked, it would be trivial to perform the next step, pass it on to a calculator and return the right answer.
> I recently asked gpt to give me a markdown table of the continents, their landmass and their populations. I checked the numbers and they matched wikipedia.
But there is no reason to think that it will be correct, structurally. While you can find examples that work, the AI does not know the truth. An AI does NOT have an understanding of populations, locations, etc.
Google didn’t return answers, but it directed you to sites with answers. Those sites were like curated by hand by a human who does understand the meaning of the data
No tech invention ever kicked out the previous iteration. Google and ChatGPT will co-exist. Same way faxes are still alive even if mobile telephony conquered the world.
What will happen is that new use cases will be set in place.
Some people might head to Google for diversity of results. Others will prefer ChatGPT for one single search result.
In my opinion, only the "I'm feeling lucky" function on Google will be challenged or replaced.
How can a tech journalist imagine that noobs will modify the default search engine on their Chrome browser and Android smartphones? These guys are not realistic.
The point made by the article is that these big lawsuits distracted management. And Google's management is so ineffective today, imagine how much worse it would be with this crowbar poked in.
> But Alphabet will probably be fine. YouTube and Android are still...
If you look at a revenue breakdown[1] and then also figure in that search has much higher margins than the other areas, it's pretty skewed to search dollars. They've improved it some, but there's a long way to go.
YouTube being a billion user product has nothing to do with how successful they are. My free service of "Contact me and I will give you a dollar" can grow to billions of users if you borrow enough money, and don't care at all about actually making it back.
AI is not the issue here. There is this very large assumption that just so happens to underpin a VERY large part of the tech industry, not just Google. It is assumed that ad spending will always be worth it, ad revenue will always increase, and that the market for digital advertising and targeted digital advertising in particular will always be large enough to support the behemoth that it is today.
But what happens when marketing spend goes down (as it does in a recession)? What happens when customers aren't willing to pay as much for your advertising product (as happens in a recession)? What happens when borrowed money can't be secured easily, and debts come due (once again, as happens in a recession)?
Are people really ready for an economy where user data isn't massively lucrative? Are people really ready for an economy where advertising doesn't pay enough to support the "free" services we take for granted? Are people really ready for a world where free video hosting, free hard drives, free email, free cloud-run office software are not sustainable?
> It is assumed that ad spending will always be worth it
This is mostly true. If it wasn't worth it, businesses wouldn't spend money on it. At my last startup we spent money on ads that were worth it, for us, so we stopped. We also found ads that were worth it, so we spent more.
> ad revenue will always increase
Ad revenue is cyclical, it's often the first thing to get cut and the first thing to come back, but it's always going to be there in some form barring a completely new economy.
> the market for digital advertising and targeted digital advertising in particular will always be large enough to support the behemoth that it is today.
Advertising shifts around from time to time (e.g. print->TV->online), but it seems unlikely that it's not going to be digital for the foreseeable future, and targeted will likely always outperform non-targeted. That's not to say it will stay at any particular company/property/app/etc. but unfortunately at this stage it's probably only going to concentrate and migrate from one behemoth to another like it has for most of the history of the industry.
It's OK if marketing spend goes down as long as digital continues to steal from non-digital advertising. Example: Advertising money spent on barely-targeted billboards and TV commercials could move to digital substitutes. This is some of what Google is trying to do with YouTube purchasing the rights for some NFL games.
Digital can steal all they want but if the thing you have attached to the ad ends up costing more than the ad gives you, market share means nothing. All this means is you lose even more money.
> But what happens when marketing spend goes down (as it does in a recession)? What happens when customers aren't willing to pay as much for your advertising product (as happens in a recession)? What happens when borrowed money can't be secured easily, and debts come due (once again, as happens in a recession)?
Typically - what happens is - central banks lower interest rates until they're negative in real yields - and people start spending like crazy again.
Who knows what will happen in the future. But that's been the previous playbook.
that only works if people A) still trust the underlying currency and B) you don't create hyperinflation. What happens in the future when the global economy isn't denominated in petrodollars?
Thats because the rest of the world need US dollars to buy fuel. We and as such we have the defacto world currency. But that is changing, every years green renewable energy makes up a higher proportion world energy and we use less carbon based fuel requireing out currency less and less. Secondly we are seeing other world powers build out the infrastructure for international banking denominated in other currencies to avoid the US banking regulation and sanctions. at some point the Dollar wont have the level of utility and dominance that to allows the US to do that sort of currency manipulation without consequences.
Weren't all these questions answered during the great recession, It wasn't easy but Google weathered the storm. Why is any upcoming recession going to be substantially different.
Maybe a part of Google's problem is it's focus on "personalized" advertising.
Simple, context based advertising is much easier/simpler to implement --- no privacy invasion required. It also annoys the consumer a lot less and by some accounts is more cost effective.
Amazon in a ecommerce site. If I go there and search for widgets, I should reasonably expect (and maybe even appreciate) some ads from widget makers. What I don't expect and what I find annoying is ads for widgets following me all over the internet for days; even on non-commerce sites; even after I already made my widget purchase at a brick and mortar --- aka, the Google "personalized" approach.
"Personalized" ads are really kinda dumb --- but apparently Google has "sold" enough advertisers on the idea so they can charge a premium for them.
I don’t understand arguments that say google will be left in the dust when it comes to AI. There’s no network effects like Facebook, it just takes a lot of money to train the models, which google obviously has. Do we really believe google, with all its resources and engineering talent, has no ability to develop something on par with OpenAI?
They also already have best in class AI. Google's PaLM is a much larger model than GPT3 and even larger than most estimates of GPT4. They also have Imagen which blows the doors off of Dall-e 2. In no way is google behind OpenAI and they have access to WAAAY more training data and resources in general. Google just hasn't shipped a consumer or publicly available product with any of it yet.
> Do we really believe google, with all its resources and engineering talent, has no ability to develop something on par with OpenAI?
Yes, they’re a large, bureaucratic organization, that’s incapable of developing new products or avenues of growth.
For the sake of argument though, let’s say Google created a product on par with ChatGPT, they would just kill it after a few months like every other Google service that they’ve made.
In fact Google has a bunch of SOTA LLM models, some of them are (conservatively speaking) several months ahead of OpenAI's. It's really interesting to see those people who believe (or want to see) Google being an underdog in the area of machine learning, but it's actually not. It's something similar to say that Tesla is years ahead of Waymo or Cruise.
Google acquired DeepMind almost a decade ago. I'm not sure if anything significant has been produced yet? Maybe it has, and it's all behind the scenes.
I don’t think comparing IBM to Microsoft is fair or accurate. Microsoft is incredibly relevant and valuable today, many years after ending its antitrust investigation.
Google hasn't been a trendsetter for a few years now, either (unless you count enforcing web standards with Chrome). Also, calling "relevant industry giant making heaps of money" a failure scenario is quite a stretch.
But ChatGPT is widely considered as an MS product, behind the OpenAI branding. MS under Satya Nadella is a different company from under Steve Balmer; it is setting, or at least trying to set the tech trend.
People who complain about Google remind me of the meme “no one goes there anymore - it’s too full.”
Google has a lot of legitimate headwinds, but honestly I don’t think antitrust or AI competitors are them.
The legitimate “destruction” of legitimacy on the internet, thus making Search more error prone is an issue. Closed off ecosystems such as apple is an issue.
The antitrust case is a big deal, but it’s unlikely Google would be broken up due to it.
I made this comment in another thread but I can't decide if LLM / GPT is a seismic shift in search or just a gimmick.
I remember when Copilot came out last year and everyone was up in arms about how it will replace most developers, then the chatter quickly died when people realized that it's not that great outside a few specific scenarios, it just looked like it could kill the SWE industry from people that haven't been inside of it.
Now it seems the same is happening with ChatGPT. While it's easy to draw conclusions based on what you've seen ChatGPT do, I wonder if it will provide long term value or be a gimmick much like Copilot, having a bunch of simple applications but fail at the more advanced applications (which would be required for it to dethrone say Google Search).
Personally I think LLM/GPT will never replace the current Google search, it however will be baked into almost every search engine and allow an "alternative" way to search the web, where you get answers to simple questions right away without having to go to a website.
> it however will be baked into almost every search engine and allow an "alternative" way to search the web, where you get answers to simple questions right away without having to go to a website.
That's what Google does quite a lot already anyway. The difference is that Google directly cites its source and offers alternative links - which is necessary, since if you need to verify the answer with a Google search, you might as well skip asking ChatGPT (or an equivalent) beforehand.
Since ChatGPT is fundamentally trained to do a different thing (word completion), I have a hard time imagining that you can sufficiently coax it into replacing an actual search. But we will see.
I think this is the author's wishful thinking more than anything else. Google isn't going to face some existential threat by Microsoft+OpenAI, at least not in the near term because the biggest chunk of the work to get a good AI response in a search product is retrieving good context to feed it.
How do you do that? Well you need an index of the web and a good ranking algo... oh wait.
Sometimes, I wonder if for the sake of small business and general good, if horizontal integration should be heavily regulated and restricted for publicly traded companies (not private ones).
For example, Apple. Why aren't the software and hardware divisions separated, at a minimum? Furthermore, I might wonder, why couldn't iPhone, iPad, Mac, Audio Accessories etc. be separate companies? Sure, you'd lose a ton of the integration that makes it special... but I wonder sometimes if the world would be a better place. Same for Microsoft: Why can't Windows, Office, Azure, Xbox, all be separate companies? If Windows was an independent company I think it would probably be way healthier for Windows at this point (less Microsoft bloat)...
I feel (just feel) that publicly-traded companies should be forced to pick a market segment and stick to it, and any other businesses that get too large (say, $1B revenue) must be broken off.
Here's the thing: That might be true, and yet, necessary for the good of society and the good of the long-term.
See, for example, Standard Oil. They massively lowered oil prices - we broke them up. See also AT&T. Once their usefulness as a monopoly ran out, we broke them up.
Plus, we can see that hardware-software coevolution has large benefits - but who is to say that having companies separate does not also have large benefits of its own? You might not have Windows shoving Microsoft Teams and Edge and Bing down your throat... or Apple pushing Apple Music to your parents. If Apple can't integrate, perhaps we'd have more open standards for interoperability.
Regulating horizontal integration seems like it would be very difficult. How exactly would those horizons be clearly defined?
I’ve always thought of Apple having an emphasis on vertical integration(hardware+software=product). So maybe their foray into services would be the line in the sand?
I tend to agree with you, or at least the sentiment of “not perpetual growth”, but I think then we run into the problem that the US might lose dominance in various industries to countries who won’t throttle their corps like that
Which maybe isn’t the worst. Maybe we’d see an exodus of the rats in our own govt and leadership as they chase their efforts into the new world leaders.
Maybe this is the end stage of capitalism and while it was a good experiment to get us into worth surplus it comes at the cost of liberty. Maybe at some point we just need to tweak the entire economic system into the next era.
Why can't Google compete in AI? They have major AI researchers with legendary experience under their belts, what is stopping Google from building something like ChatGPT to augment their search? I guess I don't understand where the threat is. They have a lot more resources than OpenAI.
Yes, because this is a solvable problem probably easier than figuring out how to make it a useful product (like knowing what it doesn’t know and preventing convincing hallucinations in some scenarios while maintaining the ability to hallucinate for creative queries). There is a lot of work on approaches like using search retrieval methods like embedding similarity lookups, and then pasting summaries of the retrieved documents in as context. This allows the LLM to cite new data without retraining. (That’s also something google is very good at)
Can somebody explain to me the existential threat AI is posing to the search business?
I am familiar with ChatGPT composing reasonable sounding answers to prompts. I get that I could type in "what's the best vacuum cleaner under $500?" or something like that, and both Google and ChatGPT could generate an answer for me.
But I want the best, most accurate answer, not a set of banalities that a neural network has composed to convincingly pass the sniff test. I actually care about the content being right, or as close to right as possible.
I admit freely that I'm clearly missing some possibilities here. Can anyone fill me in?
I'll give it go – Essentially the quality of the average website is declining due to poor incentives such as SEO optimization and content creation farms. Meanwhile, language models like ChatGPT are becoming more advanced. This presents a problem for Google, as ChatGPT will over time become more likely to provide accurate answers than a website search. Additionally, many websites will begin using ChatGPT to generate large amounts of content, making search even worse and more efficient to ask ChatGPT directly.
I think the opposite: Google will do great. Google+Deepmind's ML research groups are leading everyone. While it's true that they are late in getting something like ChatGPT to market, I think it was a wakeup call and it will be good for their ad revenues once they do. When they release it, they'll know much more about users' intent than from a simple search engine query and users will stay on their properties for longer. The concerns about cost are overblown - they don't need to run it for every query.
Google is a public company so its shareholders want the stock price to be higher ASAP. Being an unquestionable leader in AI with a product as good as ChatGPT (and maybe an image model available to the public) would help in that goal.
I really don’t think it would. Google is already number #1 in search. OpenAI is quoting a ~$40 price per month for ChatGPT Pro right? That’s not an insignificant cost for 99% of the population. I don’t see how it makes sense for Google to release something until they actually need to.
I tried to imagine an ad in chat output by asking ChatGPT a question and then making up an ad:
Q: What is superglue?
A: Super glue, also known as cyanoacrylate, is a type of adhesive that forms a strong bond quickly and can be used on a variety of surfaces. It is often used for small repairs, such as broken ceramics or toys, and can also be used in crafts and other projects. Super glue is available in liquid or gel form, and is sold in small tubes or bottles. It should be used with caution, as it can bond skin together and can be difficult to remove.
For your superglue needs, you can't do much better than Gorilla brand superglue. Its high strength and quick set time make Gorilla Super Glue the go-to adhesive for a variety of household projects.
Edit: There used to be a radio guy, I believe it was Joe Franklin who blended the ads into his spiel so you couldn't tell when the ad started and left off, so maybe it will be something like this.
A: Super glue, also known as cyanoacrylate, is a type of adhesive that forms a strong bond quickly and can be used on a variety of surfaces. A brand like Gorilla with it's high strength and quick setting time making it perfect for for small repairs, such as broken ceramics or toys, and can also be used in crafts and other projects. Super glue is available in liquid or gel form, and is sold in small tubes or bottles. It should be used with caution, as it can bond skin together and can be difficult to remove.
Prompt: Write some ad copy for me in the style of a friend recommending I use superglue
"Hey friend, have you ever had a broken vase or a pair of glasses that you thought were beyond repair? Well, let me tell you, super glue is a game changer. It's strong, fast-drying, and can fix just about anything. Trust me, I've used it on everything from ceramic to plastic and it works like a charm. So, next time you have a household emergency, don't throw it away, just grab the super glue and fix it up like new. You'll be amazed at how easy and effective it is!"
Prompt: Write some ad copy for me in the style of a friend recommending I use Gorilla brand superglue
"Hey there! Are you tired of using weak, unreliable glues for your DIY projects and household repairs? Look no further than Gorilla super glue! This stuff is seriously strong and durable. I've used it to fix everything from a broken toy to a wobbly chair, and it holds up like a champ. Plus, it dries quickly and forms a bond that's almost impossible to break. I trust Gorilla super glue so much that I wouldn't use anything else. Give it a try, and you'll see the difference for yourself."
I remembered it wasn't Joe Franklin, it was Paul Harvey that blended ads and content. I don't recall him giving any disclosures which is why I didn't like his radio talk show.
Perhaps the author has simply chosen to discard cognitive structures that will be made redundant in the new post-literacy age of ubiquitous ML language tools. Would that we all had such vision, and the courage to act on it.
Google in 1998: Appendix A: Advertising and Mixed Motives from the original paper The Anatomy of a Large-Scale Hypertextual Web Search Engine [0]
"Currently, the predominant business model for commercial search engines is advertising. The goals of the advertising business model do not always correspond to providing quality search to users... we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers... we believe the issue of advertising causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm."
Google in 2003: Eric "You can trust us with your data" Schmidt convinces Sergey and Larry of the bonkers amount of money they can make by adding a monopoly on digital search advertising to leverage their monopoly on digital search. They proceed to switch to the same advertising business model that hindered their early competitors.
Google in 2008: Buys out Doubleclick to maintain Adsense's monopoly on digital search advertising and doubles down on surveillance capitalism.
Google in 2011: Buys out AdMeld, changes AdX contracts to be even more anticompetitive to maintain monopoly on digital search advertising.
Google in 2017: Loses $2.7 billion antitrust lawsuit in the EU. Response? No changes to business model, but oust a vocal supporter of the decision from Google-funded think tank New America. [1]
All big corporations have to be broken up. They have ill influence on everything and they are what many describe symptoms of late stage capitalism and our failure to regulate.
They are like a player of Monopoly who bought everything.
Alphabet shouldn't exist in its current form and so Apple, Amazon and many other corporations.
Truly, the problem is US complacency. It's a great day to tax businesses when US companies dominate the market, but as soon as a foreign competitor steps in? It's time to defend American competition and ingenuity!
Hopefully, pushback from European regulators who are harder to fool (or buy out) will force the US to admit there's egg on our face. We've let corporate America write the rules for too long, and our democratic control over their abuse has become dilute and helpless. Something's gotta give, and I'm cheering for the countries that take user privacy and consumer rights seriously.
> Hopefully, pushback from European regulators who are harder to fool (or buy out) will force the US to admit there's egg on our face. We've let corporate America write the rules for too long, and our democratic control over their abuse has become dilute and helpless. Something's gotta give, and I'm cheering for the countries that take user privacy and consumer rights seriously.
Hopefully. But don't underestimate the general public's abject apathy and religious belief that more "freedom" in markets fixes everything, even when it is obvious that bad actors are employing dark patterns to exploit the anarchy.
One downside to a bunch of small companies instead of a few large ones: competition is fiercer, the small companies can't rest on their laurels the way the large ones do. Consequently work life balance deteriorates for the average worker.
How is a "big corporation" defined? Limiting the number of employees a company has (1000? 10,000? 100,000?) would have interesting effects. Companies might be less likely to hire junior people. Bigger companies would need to outsource or hire contractors to avoid hitting their employee limit. There would be competition among companies providing outsourcing or contractor services. That could be good for companies outsourcing work (having more choices), but there would be a lot of market inefficiencies.
How big is big? And do you not want computers that can do what modern computers can? It costs billions of dollars to build a single semiconductor fab, so without any big corporations, there won't be any new fabs (or anyone big enough to own an existing one), so we'll be going back to 70s-80s IC technology at best.
Like it or not, a lot of things require very large companies to get done. Small companies can't build jumbo-jet airplanes.
The big corporations that are doing the most harm are the food companies. They inject animals with antibiotics, kill millions of them in a very cruel manner, employ cheap illegal labor to process it and contaminate the environment beyond repair. The tech industry is bad but nothing compared to the irreparable harm the food industry causes.
Furthermore, the product that made their name is thoroughly lost. I dread Google Search today. If I found something on Google yesterday, I have no idea if I'll be able to locate it through the sea of blog SPAM and SEO garbage tomorrow. Google web search strongly favors big websites when sometimes the page you need is never going to be there. As an example, Google puts heavy weight on near-SPAM sites like Pinterest which produce almost no original content, and on abusive hosts like Fandom (formerly Wikia) who manages to beat out community wikis by replacing admins that try to migrate off and winning in SEO.
I am not trying to suggest Duck Duck Go or Bing are actually better overall, but if I see a web search covered in results like that, I sometimes find that they do in fact give me somewhat better result rankings.
The only thing I really have any strong affection for is YouTube, and the people and content that made the platform special are constantly unhappy with it's corporate direction, not to mention the utter lack of support, claiming and affirming in appeals stupid pointless crap like "you said a naughty word in the first ten seconds of the video" when anyone with working ears can easily verify they did not.
And that's the tip of the iceberg. I think a lot of ventures, like GMail, only ever made sense as computers and bandwidth just kept getting dramatically cheaper. It's difficult to beat free. Well, what happens when those things stop getting cheaper, and GMail keeps getting more expensive to run?