Sorry to break it to you folks but if you hated your data being with Google, you are in for a surprise on how atrociously bad Microsoft is at privacy and security. Multiple close friends who have worked at Microsoft tell me that search history data - who is searching what, is basically sitting in systems with ACLs so bad that 20-30K employees have access through transitive membership of groups. To access a customer's data you just need to know token which is logged everywhere and is apparently very easy to generate. I have heard horror stories from them about privacy incidents which never went public.
From what I know about Google, they are serious about least privilege type of stuff internally and employees dont get arbitrary unbound access to systems or data.
> Both are evil but I agree that Google is less evil than Microsoft.
I'd agree they are both evil in that they both seek to leverage your personal data against you for their own gain, but it sounds like Google does it while being more protective of what they've taken from you while MS also leaves you vulnerable in new ways through carelessness.
I'm not too worried about Bing though. That's nothing compared to Microsoft's access to your computer at the OS level. We know anything we do online will be seen by others, but being able to snoop on our personal files and log our keystrokes is a whole lot worse than knowing what we type into a search engine.
However improved/more popular Bing becomes, eventually Bing will end up in the same position Google is in now where they'll have to make their web search worse in order to constantly shove ads in your face instead of returning results that are useful, and slowly SEO spam will adapt to pollute Bing's results further. Google may not even mind if Bing starts gaining users. At this point Google can collect so much data from android devices that I doubt they need google search to peer into our lives like they used to anyway.
> Azure is the worst of the 3 cloud providers. Horrible developer experience and documentation and reliability
Don't forget security! Multiple serious and extremely trivial cross-tenant security exploits only in the past couple of years. AWS and GCP have had none ever, the much smaller OCI has had one.
Both are driven by a profit motive, but Google takes some forms of privacy more seriously.
Lets not use emotionally driven, ambiguous terms like 'evil' to describe these kinds of things. Google is not, for instance, conspiring to commit genocides, nor is Microsoft planning a coup in an Eastern European nation with the goal of colonizing them to use as indentured technical support.
It's a slightly different from of eeeevil, up to you to decide which is lesser, or if the distinction has any meaning at all.
Microsoft is the capitalist evil. Google is the technocratic AI evil. The first wants your money, as much of it as they can hoover, using dark patterns if useful for it. The second wants to optimize your life using _their_ metrics, without ever engaging with you (humanity is the problem - engaging with it never scales).
I wouldn't be extraordinarily surprised if this is the case, but I do not expect it to be as lax as you suggest and given how it's presented to us it reads like someone trying to get people to not look at something. If you have things you can present as evidence that would carry water for your argument, otherwise it's random inflammatory claim on in the internet.
And I agree, Google does take security more seriously than most places.
Of the major three cloud providers, Azure is by far the one whose security bugs I hear most about. Not that this necessarily has much relevance to search, but it doesn't inspire confidence in their culture.
Oh yeah IIRC I used to look at Bing search logs in Microsoft and I wasn't on Bing team. This was 2014. Things probably have changed.
IIRC there was some basic safeguards though, like a query must have had like 10 unique occurrences or something. I also have no idea if you could tie searches to people.
I was just looking at aggregate numbers. That was really cool. It was like Google Trends but with real numbers!
I hate Big Data(tm) as the next guy in line, but if I have to choose between Google and Microsoft nomming my data I'll happily pick Microsoft every time.
Why?
For one, I've had a Microsoft account for over 20 years now (anyone remember Hotmail?). It's long past due for me to be complaining, and Microsoft hasn't wronged me in that time anyway. By comparison, my oldest Google account only goes back just over 10 years, and horror stories abound even if I've been fortunate so far. I keep all my truly important correspondance and login tie-ins with Microsoft (read: my Hotmail).
For another, Microsoft nurtured over 30 years' worth of good will from me with Windows and Office; even though I hate many things about Windows from 8 and up, among other things, I will ultimately be a friend to Microsoft simply because they were a significant and positive part of my childhood and now my adult life.
> Microsoft account for over 20 years now (anyone remember Hotmail?)
From memory, Microsoft used to be absolute bastards when it came to your data. They did everything they could to keep you gated in and without any control. Things like: refusing any export tools for your emails so you couldn't migrate to another provider; or, not allowing any non-web access to your email.
Then Google came along with GMail and revolutionised things: amazingly generous free data allowance, IMAP access, data download options, open standards, liberal approach to letting people set up various hacky app access, and so on. It was hugely successful for them, and forced MS to moderate their approach.
Google abandoned their Don't Be Evil strategy and sacrificed a lot of their principles (and, I would argue, a lot of their success with it), but at one point they had a positive impact and they've never been as evil as early-era Microsoft.
Dunno if it was incompetence or malice, but the straw that finally broke me over to Linux permanently was a privacy setting in Windows that appeared to work until you re-entered the menu and it wouldn't persist. In my time at YouTube though, privacy issues were an immediately fireable offense, and even when my role required data, use was gated and then still logged, and misuse was handled. That's an n=1 apples-to-oranges data point, but I personally trust Google a bit more than MS with my data.
I don’t hate Google for having my data because they have worse internal controls.
The monopolistic dominance is the issue. They have too much data and too much power, and leverage that to swallow other market categories. Today they have a deeply unhealthy dominance over the tech industry.
Google could do everything right and still be an existential problem. You say MS has data security issues, okay sure. I’d still rather have them or anyone at 25%+ of the market.
Apple isn't a monopoly yet it has huge power and domnance with safari. Do you avoid Android because it's the most common or Linux since it's got a monopoly on servers?
Google isn't forcing someone to use it. It's monopoly can disappear when something better it's there.
That anecdote reflects the data protection part, everyone dealing with Windows trying to disable tracking experienced their horrible data collection attitude
It's not much about Microsoft. At least it improved human life in serious ways (Typescript, VSCode,...). What did Google bring to my life ? Nothing except for spam stuffs from Ad, Email,...
You're responsible for your data, not Google, not Microsoft.
Uh... Maps? Open, Permissive phone OS compared to the alternative? Greatest entertainment/information system, especially one that's democraticsed producing media?
In terms dev-related stuff, TensorFlow - literally something that's made today's ML possible, Kubernetes, BigTable, MapReduce, and yknow, that good old thing called Chrome..
I understand hating on Google. They're not that great, but none of the FAANG are "great".. They all have some terrible bits, some good bits. However, to say that Google has not provided anything except "spam stuffs" is very wrong.
Search alone has changed the entire way that humans consume information. I would say that today's internet landscape, for better or worse, would not exist without Google.
While it might be fashionable to hate on Google, the question "What did Google bring to my life" seems really odd. I can't believe I have to say this but search, and more importantly good/fast search, is to Google's credit. Typescript and VSCode all have, and will continue to have alternatives, but for a long time, Google was (some may argue it still is) the bleeding edge of search.
Anyway, there might be a generational thing at play here, if so, I meant no condescension.
What an absurd statement. I don't have especially strong feelings for Google in any direction, but I can't deny that, even if Google had only done search or Maps, that alone would be a massive boon for society.
The article starts, "Nestled in the hills of Mountain View", which is worrisome, because as anyone with even a passing familiarity with Silicon Valley geography knows, Mountain View is mostly flat. And Google's Mountain View offices are located on reclaimed baylands, which are especially flat.
I wonder if the article authors were thinking of Xerox PARC's offices in Palo Alto's hills? Or the road named "Sand Hill Road" that used to have some venture capitalist offices? Other than those offices I can't think of any significant hill-based offices in Silicon Valley.
Although Google, Maps, Youtube are of daily use they are monetized exclusively by advertising which is annoying and hated by many.
It has been many years since Google has launched an innovative new product.
I don't think ChatGPT will gain daily traction after this hype. Anyway we could say that MSFT and AMZN have demostrated more power to innovate with different business models (not only adv) and products.
The GOOG stock has a PE (Price/earning) of 23, while Coca cola 26. So the stock market expect higher growth from CocaCola than Google. Quite surprising.
Google stopped being innovative the second their "20% Time" got gutted:
> Employees were encouraged to spend up to twenty-percent of their paid work time pursuing personal projects. The objective of the program was to inspire innovation in participating employees and ultimately increase company potential. For Google's part, Gmail and AdSense both arose out of side projects. In 2013, Google discontinued 20 percent time.
What has Google released since 2013[0]? Stadia?
Then you tack on that Google Search as of today is basically unusable/broken with people needing to append site:, double-quoting everything, or enabling verbatim search to get high quality results. Heck, on top of that, I also have a uBlock configuration that removes 280 domains from Google Results because they're all auto-generated spam of Stackoverflow answers.
There is something deeply wrong with the culture inside Google.
>I also have a uBlock configuration that removes 280 domains from Google Results because they're all auto-generated spam of Stackoverflow answers.
Can you describe exactly what you mean? If there is a way to hide specific domains from showing in search results, it would be an incredible boon for me.
Some have postulated that this might be Google's Blackberry moment. On the flip side I wonder if its actually more like its Dropbox moment.
The press loves a David and Goliath story; the young disruptor versus the stodgy disrupted.[1] At that time, Dropbox was David, coming out of nowhere in a hurry and had a product that seemingly should be so obvious for Google to launch but hadn't - making Google seem antiquated and slow.
Fast forward a couple of years (aka a decade) and Dropbox is still going ok but Google Drive is by far more ubiquitous.
Google makes fuck all revenue from Drive, it never mattered. EVERYTHING at Google, most notably the cash fire that is Cloud, is funded by Search. This is the first time that is in danger and over 25 years they never came up with a fallback plan. This is an unmitigated disaster and purely the result of mismanagement.
They didn’t exactly stop innovating: ChatGPT itself is using tremendous amounts of innovation that was done and published by Google. What they stopped is beating their competition to launching compelling products based on the innovation they do.
IMO this whole narrative reflects a type of human bias and the reality that we didn't evolve to reason about trillion dollar complex business.
I love chatGPT but it clearly has a ton of issues. To declare Google dead that is point is completely absurd.
I can't think of a single transformative technology that actually has a first mover advantage. The first, second and third mover seem to make all the mistakes and setup the forth mover to learn from those mistakes for free and create a monopoly.
Counting out a company that owns Deepmind is just not rational.
Nearly all of the authors left Google years ago. If the company doesn’t keep the people innovating it’s not like they’re some magic middle manager who can make it happen without them.
I think YouTube, which has a paid and ad-free version along with its Live TV offering, somehow manages to still be a sleeper hit within the larger Google portfolio and doesn't get talked about enough.
What happened to the Waymo hype? If Waymo manages to break through and become a sufficient AI taxi, then car ownership becomes optional for a majority of Americans (unlike today).
Ads aside, the UI of Youtube and Maps is atrocious. There is so much room for improvement here, but it will never happen from Google. It's just not in their DNA. Unfortunately, they have such a huge moat here I doubt we'll see anything better for a long time.
They used to have pretty good UI for maps until some PM needed to make their mark... Given they count individual pixels and +/-1 deltas in colors I don't think they are ever going to get any UI correctly. The initial simple UI with just a single search input box that was highly praised by minimalists was just an accident.
Generated playlists (recomendations) are always (for days/weeks) the same. Not a bit randomization. It feels like they were approved by party commitete for years.
I don't understand why they decided social network in music service.
Ah hell. CUE-lists are still too novice technology for them.
I still think that last.fm 15 years ago was the best music service.
If search gets usurped, or even a marginal but meaningful reduction in usage due to competition from AIs, then its not surprising at all that Google’s near term prospects could be worse than Coke.
Even Google winning the AI wars leaves them worse off as the operating profit potential of querying an AI vs their search index has to be far lower.
As to usage, once you start to integrate ChatGPT into your workflows it can meaningfully benefit vs traditional search. Ive been able to find information on specific programming language concepts, with generated examples, far faster than searching.
I can ask it about GameDev concepts and ask for bulleted lists or higher/lower level of detail in the answers. Information is presented in a much easier to consume manner
That being said, most stocks that are considered defensive are quite overvalued on a fundamental basis. I would consider a PE of 26 for Coke quite undesirable, though there’s much worse
If Google ends up winning in the AI war, search and ads will be significantly cannibalized, but it will have tremendous opportunities outside of search, many of them in areas that don't even exist yet.
In some ways, it might even be good. It's suffering from a kind of advertising resource curse nowadays, and being forced and able to diversify its economy would be best in the long term.
>It has been many years since Google has launched an innovative new product.
I can say the exact same thing about countless other large companies: Facebook, Intel, Microsoft, Boeing, AMD, AT&T, I could go on and on. At some point, launching innovative new products isn't really that important for a company, when many people on the services that company provides.
Google used to be considered different from and better than all the companies you list (except Facebook). Being lumped together with those companies would have been seen as an insult to Google in the past.
It is if you don't want to wind up like IBM, i.e. - a company that could have been so much more, knew it had to change at the leadership level, and yet still couldn't get out of their own way. Apple's renaissance came because they started releasing innovative products and cannibalizing their own business.
They are kind of forced to keep up or die now, not a good place to be in. It feels like everyone always just assumed they must have super advanced AI behind closed doors. Perhaps that was a wrong assumption because they never actually seem to *do* anything with it, but we will probably know that for sure very soon.
Google are a mature corporation. The fact they managed to innovate for so long is really a huge achievement but nothing lasts forever. The same applies to the rest of FAANG one way or another. Netflix cannot innovate out of streaming tv and movies, and now face strong competition from others. Apple has always been more about marketing than real breakthrough tech innovation and the bits of real innovation they did do left with Jobs. Amazon has flopped most of it's recent attempts (again, respects for running BOTH the worlds best retail company AND AWS). Facebook has been flailing around doing crypto currency and VR and getting nowhere. Microsoft haven't innovated since Excel FFS. Coke at least has the occasional new flavour...
I'm not "blaming" them, it's the cycle of life. Gravity eventually wins.
This is also a key reason for all the layoffs: Growing, innovating companies cannot ONBOARD staff fast enough, because their value is decided by how big they will be in the future and how fast they can get there. People are a profit centre. But mature companies' (in mature industries) only source of increased value is in efficiency, getting the same revenues with less costs. So people are a cost centre and they succeed by showing they're efficient (aka laying off workers).
It is depressing that people dont recognize that Google has invented and perfected a parasitic business model that has destroyed the internet (and much of tech with it) and this episode feels like just another turn of the downward spiral.
The article aims to offer a preliminary analysis of whether Microsoft can become a better parasite: Grabbing the content people generate online, paying nothing, and using it deftly to serve advertisement on the basis of private behavioral traits that are gleaned by prying open and subverting the use of all pieces of IT people use.
It is not too difficult by now to imagine alternate tech universes (Philip K Dick style) that have nothing to do with this nightmare, where more or less the same technologies empower individuals and companies and organizations rather than squeezing them dry. The combination of oligopoly status, moral laxity and political dysfunction means we are simply sitting around like sheep discussing whether a new butcher is about to get sharper knives.
I grew up using the internet as a child and when Google came out it was way more a blessing than a parasite.
The web was hard to navigate and you relied on webpages "befriending" each other and helping you navigate to similar pages. That was real navigation and it had terrible recall.
Once google became the Internet's frontend everything was one or two hops away. It was awesome, and I think that it has evolved in a way that saves time with the snippets and knowledge graph panel at the expense of taking away clicks from sites that needed to serve their ads. Although today things are getting worse with walled gardens and SEO.
A problem that the internet has not solved yet is how to keep webpages up while there's increasingly more content. Most sites were maintained at a loss by people, or leeching resources at a university or company. The closest "people" have got to fix this IMO is through P2P, but it got too focused on piracy and got a bad reputation with malware too. And I think we never had the infrastructure to serve and share webpages we visited. Sharing today is still a mess and we still depend on centralized distribution + caching (and that's dying with https, only big companies who are better off giving caching servers away to ISPs can do this).
Ads came to "save" the internet from this problem. I really hate them and I wish we had a better model, but they seem a necessary evil as people just got used to believe that all digital things are free when they obviously cost something, and that still shows today in the apps stores.
> Ads came to "save" the internet from this problem. I really hate them and I wish we had a better model, but they seem a necessary evil as people just got used to believe that all digital things are free when they obviously cost something, and that still shows today in the apps stores.
Advertising worked just fine for more than a century, ever since the invention of consumerist society. (How much of that is really necessary in a society that is pushing headlong into unsustainability is another, bigger story.)
The disaster we have now in our hands is of a more immediate and crass nature.
Think about it: there are algorithmic breakthroughs that can help everybody on the planet raise their level of information retrieval and the only discussion is which of the former or the current consumer tech oligopolies will use it best to push ads and how to double down on a bad design.
I would have agreed with you in the past. Now, I have to disagree.
>the web was hard to navigate and you relied on webpages "befriending" each other and helping you navigate to similar pages. That was real navigation and it had terrible recall.
This is exactly why internet was decentralized. A network of social connections between small online spaces people congregated in was essential to find what you wanted. It was somewhat of a replication of the old world - where in order to find an answer, you first had to get into relevant circles (eg. a local club, university, etc). You didn't just get an answer to your question, you inevitably had to sift through a lot of other content and learn about new places.
Google cut through all of that. Ideally, you could just get the answer you wanted, and not even see the rest of the site. The first order effect was that everyone's life get better. The second order effect is that it killed the old decentralized web, because random discovery nearly died down.
LLM are only the next step on this, but I don't think they change much. I think it's mostly going to damage reddit. A lot of reddit questions are an attempt to find an answer to something that google is too dumb to find - but llms may be able to.
The defense to this - are chatrooms, like discord. Sometimes I see people complaining that so much information is now 'hidden' on discord. But that's the exact point. Making information hard to get means people are forced to interact with each other. This creates incentives to contribute.
That's the future. Chatrooms may be replaced by voice based vr, hard to say. We already passed the peak of public information dissemination, and are going back to old style "ask at the university", just more decentralized and online.
Don't tell me you forgot Yahoo, Excite, Altavista, Lycos, and all those other search engines existed! Lycos also had hotbot and tripod.com, back when homepages were your internet presence instead of social media.
Maybe Google had results that were a bit more relevant and won that competition, but they didn't Change Everything.
> It is depressing that people dont recognize that Google has invented and perfected a parasitic business model ... The article aims to offer a preliminary analysis of whether Microsoft can become a better parasite
In this model, are end users who use it to search for information parasites as well? After all, they are consuming content that other people generate online, and usually pay nothing for it. And they love to look for — and find — what they are interested in. If Google didn't satisfy that need well, people wouldn't have preferred it, and it wouldn't have grown.
Users pay with their data. That is the whole point of the business model and it is obviously lucrative enough so that the entire tech infrastructure (devices, OS, browsers etc) can be repurposed to be a user data collection channel. If you want to find a real accomplice that is essential for the model to work, it is not the users, it is the advertisers. It takes two to tango in the adtech market.
I don't dispute that search (in its various incarnations) is an essential service in a digitally interconnected world. There are countless ways to pay for it (as a digital public good, as a user subscription etc) that are fundamentally better than what we have. It is also obviously true that some decades ago Google innovated technically. A lot happened since and it wasn't positive. Normalizing it simply prolongs the agony.
Google started out noble, but was poisoned by its own success and by extension, poisoned everything it touched. The web, as a platform, is nearly dead. Walled garden phone apps are almost completely dominate. The revenue model is on such shaky ground that almost all remaining websites either went clickbait with a pox of ads or paywall. I would not be able to list more than 5 websites that survived Ads and SEO, perhaps only Wikipedia.
These LLMs will be the end. They will capture enough of the user attention that the few remaining ad supported sites will be wiped out. “Content”, that low quality grist every site is filled with to capture users for a few seconds before they click on will become worthless, no different than the low quality crap the LLMs spew out.
Finally, we will stop trying to keep the bloated corpse of the web alive and will move on. It will be the death of a dream. At least we can all let go and focus our energies on a new dream. Whatever that dream is, I hope high quality, human created art, knowledge, opinion and creativity is at the center of it.
I wish I could find something to disagree. If there is any hope it ultimately comes from the fact that the technologies we talk about are at various stages of commoditization. Maybe at some point the gap between what-is and what-can-be will be so stark that it will set in motion positive forces that are currently dormant.
"sitting around like sheep discussing whether a new butcher is about to get sharper knive"
Speaking of that, I might as well provide some supprting evidence from a different part of the tech industry. Most people don't know that right now Blackrock and Coinbase are attempting a coup on the US dollar by seeking access to central bank liability for their stable coin? The Coinbase CEO even said it out loud that this year his stable coin will become the de facto central bank digital currency. I suspect that they have indirectly bribed someone at the Fed (how else does that work? esp since the Fed chairmen has come out strongly against stablecoins) and are basically trying to get the ability to get Fed loans at the Fed rate and loan out at the commercial bank rate without being a bank and by using a coin which they can mint themselves, so all the profits from interest create more reserves and the cycle keeps going. It is literally insane. The threat from Google and Microsoft to society pales in comparison. I may not have my understanding 100% correct but all banks are up in arms about it and the average person on HN has no clue this is happening. Just Google Blackrock, USDC and the Fed. The banks are calling it a backdoor CBDC.
Yes, Blackrock and Coinbase have partnered up to provide crypto trading access to Blackrock clients but how are they attempting a coup on the US dollar?
As someone else said, the problem is ChatGPT lies straight to your face, whereas at least Google's answers are based on structured data by someone. It's traceable, whereas I would not trust ChatGPT ever to tell me the correct temperature to cook a steak.
I don’t know if you saw screenshots of the Prometheus model inside Bing. But it presents the sources with all its answers. It is no ChatGPT but a new model created for search.
The SEO content served up by Google search nowadays isn't much better, and from what I've seen I don't think a lot of it is being generated by someone either.
Perhaps a bad example, since ChatGPT consistently gets the steak temperature right (or at least gives the same values as Google). Internal temperature of about 130-135°F (54-57°C) for medium rare, etc.
The correct temperature to sous vide a steak depends on the desired level of doneness. The following temperatures are guidelines for cooking steaks to different levels of doneness:
It's important to note that cooking steaks at low temperatures for extended periods of time can result in a more tender and flavorful finished product, compared to cooking steaks at high temperatures for a shorter amount of time.
Completely agree. Heck, even the current 'waiting list' page (bing.com/new) looks so outdated and a mess. I don't know whether it is the font, the colors, the composition or margins..
Bing itself looks like a mess too. Microsoft truly has no taste.
The scary real threatening and killer thing would be for chatGPT to automanage my tabs. Like a well-meaning shepherd, perhaps create new pages on the web that by a combined virtue can take the place of whole groups of open tabs, being equivalent convenient replacements of the information they once held.
As long as it's easy to see the list of web results so I can go there instead of just relying on the model. Also because there are reasons to go to sites other than just getting a summary. Like maybe you need or want to use that site to do something.
People seem to forget: AI is about data, the ability to process it, and having a few smart folks to do algorithms. The third part is actually the smallest and easiest part of the trilogy for AI.
Google, has all three. The real question is can they not shoot themselves in the foot while doing it.
Also SEO will always target the market leader.
As far as Bing's results: I keep thinking Google is trashy. Then I use Bing and I remember just how good Google is. That's my personal opinion. I am NOT going to claim they are SEO, and spam immune, far from. But they seem to do better than Bing in getting what I mean right.
ChatGPT may help MS, but is it a bandaid? Is it just a good PR bump? Who knows... Displacing Google is a large task.. and not one I'm sure I want MS to win. But I look forward to them trying, if only to make Google do better.
Google is without a doubt worse for me than ever. But is it really google? Or is it just the absolute cesspool of SEO-optimized ad driven garbage that represents a growing portion of it?
The other searches are worse. I too have tried. But in general I'm frustrated with Google now more than I've ever been.
This is really the core of my disappointment in Google: their poor handling of malicious SEO. Having humans go over shared site blocklists that people use to improve results, hand verify each entry, and remove those sites from Google results would go a long, long way.
> Then I use Bing and I remember just how good Google is.
I use a lot Ecosia (which gets results from Bing) due to them allegedly planting trees and being much more heavily into privacy.
My experience with Ecosia, thus Bing, is that for the overwhelming majority of searches (90% plus) the quality is comparable.
When you get into more niche searches Bing offers better results 1 out of 5 times and worse results 4 out of 5, with a worrying amount of those 4 being almost complete misses.
I still need to get onto Google occasionally when I don't find what I was looking for on Ecosia.
> People seem to forget: AI is about data, the ability to process it, and having a few smart folks to do algorithms. The third part is actually the smallest and easiest part of the trilogy for AI.
MS has them all three too.
And since, as you say, the algorithm part is not the most difficult part, others may enter the playing field soon (we already have DDG and Kagi, for example, with Kagi now experimenting with LLMs too).
The compute side needs to be solved with money. That may be a hurdle for DDG/Kagi.
MS has the money. Data is by far and away the hardest part, and DDG's own privacy policies etc will hurt it here, alas.
Personally, I think it is Google's race to win. But they have to DO it. If I start getting better results off DDG or Bing, I'll switch. I'm fickle, and own no loyality to Google.
I just want my StackExchange answers faster, before Google Coder, and Co-Pilot replace me. /s
I don't understand this hype and I feel like I'm looking at different products than everyone else is. There are very few complaints I have about Google that I think this technology helps solve, and for most of my complaints, getting summaries of searches makes the situation worse, not better. To be completely clear: even if the AI was perfect, I don't know that I want even an actual human being to sit down and summarize an answer to my question rather than show me a list of search results.
The problem with search is not that our answers aren't summarized well, it's that the quality of information returned for those searches is getting increasingly worse, and we are getting increasingly worse at categorizing or filtering that information in any useful way. And LLMs pulling information in and summarizing it for me is... not helpful? It's summarizing the same garbage, except now sometimes it also summarizes it wrong.
But it's not even an issue with the quality (although the quality of information from LLMs is also pretty over-hyped I think). Conceptually, I don't know that this is a product that I would ever want. I can't think of any time where I've sat down to do a search on Google or DuckDuckgo and thought, "You know what I want? I want these results presented to me in a less structured format using natural language and with less granular knowledge about where each specific statement is coming from."
At least Bing seems to be trying to do inline citations in some of its answers, which is a step up over Google's AI announcement, I guess?
Maybe I'm just in the minority on that. Users seem to like this a lot. But my ideal version of the Internet is one that decreases the number of abstractions and layers and summaries between myself and primary data rather than increasing them. My ideal Internet is a tool that makes it easier for me to actually find things, not a tool that increases the layers between me and the raw source/information that I'm looking for. I already have enough trouble needing to double-check news summaries of debates, events, and research. Getting another summary of the summaries doesn't seem helpful to me?
I can think of some ways where I might use an LLM in search, even really exciting ways where maybe it could help with categorization or grouping, but it doesn't seem like Google/Bing are interested in pursuing any of that. I look at both the Bing and Google announcements and just think, "why are you making it worse?" But who knows, maybe the actual products will sell me on the concept more.
From my point of view the only usage of LLM is generation. Such as writing peer reviews, self reviews, OKRs, when I already know the truth and can edit out any errors. I will never trust a LLM's answer to something I don't already know.
There seem to be two different directions for innovation here.
The first is a little more mundane: LLM embeddings. OpenAI currently offers an API that turns sentences into coordinates for a point in some 1536-dimensional conceptual space such that two points are close together if they are conceptually close together. This is insanely powerful. For example, you can generate captions for a bunch of images and store the embeddings for them. Then, you can look for a "picture of a rabbit eating a carrot" by turning that phrase into a 1536-dimensional point and looking for the nearest points around it. Basically, it blows open search technology for everyone. You no longer have to deal with synonyms, idiomatic phrases that mean similar things, misspellings etc - the problems you'd run into when trying to implement simple text search using traditional techniques. It all gets simplified to generating coordinates in some hyperspace and looking for nearest neighbors. This is a total game changer.
The second direction is ChatGPT. Sure, if you want to read a detailed analysis of the demographic situation in China, you'd prefer an article written by an expert. You would still use a search engine, pick a search result and do things the way you do them today. However, there's an entire collection of things that can be answered directly by ChatGPT. For example "how many mins should I hard boil an egg" or "Can I take NyQuil when I'm stoned" or anything else where you really just want a single sentence answer. Today, you launch a browser, search for what you want, skip past the first 10 advertisements, look for a site that seems reasonably reputable, click through all the GDPR warnings, scroll past the banner ads and the SEO optimizing bullshit text to find that one sentence that you wanted all along. Or, you could ask ChatGPT and get an answer instantly. (assuming chatGPT is good enough eventually).
It's hard to predict which of these two technologies will disrupt the current status quo in search. Neither might. But we haven't ever been this close to a level playing field in search since the 1990s. The excitement is hard to resist.
> The first is a little more mundane: LLM embeddings.
You know Google has been doing this for years now?
> However, there's an entire collection of things that can be answered directly by ChatGPT. For example "how many mins should I hard boil an egg" or "Can I take NyQuil when I'm stoned" or anything else where you really just want a single sentence answer.
Google has been doing this for years via search cards, which are AI generated summaries of website information.
So LLM embedding are an actual useful thing that could actually improve search. Categorization is a real problem that AI could help solve (particularly with search queries). But that's not a new category of search, it's just a question of whether the current LLMs would be better than whatever Google is currently using to make the same inferences. And the direction Bard and Bing seem to be going with these giant models is the conversational direction, and where that's concerned:
> there's an entire collection of things that can be answered directly by ChatGPT. [...] assuming chatGPT is good enough eventually
I am a lot less impressed with this. And I know I'm an outlier and plenty of people are shocked at how good GPT is at this kind of problem, so I am constantly second-guessing myself and thinking to myself, "are we using the same product?" Because I think ChatGPT produces really bad quality information. It's cool, it's wildly impressive, it's a massive achievement and an incredible milestone for AI, but 'cool' is different from 'useful.'
Leaving aside the problem that answering simple questions is a very small subset of what search is used for, and isn't on its own probably a big enough category of questions to make me change search engines, the bigger problem is that the current state of ChatGPT seems to be wildly inconsistent about what it knows and what it doesn't know and I don't have a way to pre-predict what categories of information it's safe to ask about. And the only way for me to verify the answers it gives me are to... double check its work with a real search.
I would not advise anyone to ask ChatGPT for advice about what drugs are safe to take while high, that seems profoundly unwise to me.
So it's a bit like Instant Answers. Google has been trying to auto-answer questions for ages, and in practice the only time it's ever been useful for me is when it's extremely predictable and when I know that a category of question will only ever have its answer pulled from one site and where I know what the format of that answer will be.
Unpredictability is generally a quality that I try to avoid any time that I am using a computer. One of the primary strengths of a computer to me is specificity and predictability. And so the bar here is really high. The question I ask myself is, "would I want to replace a search engine with a human assistant?" And I think the answer is no, I feel like that would be missing the point of what a search engine is. And ChatGPT gives worse answers than a human assistant would, and its sources/knowledge is just as unpredictable as a human's would be if not worse. So, I also don't want to replace my search engine with ChatGPT.
It could get more accurate in the future, and if it does then maybe my opinion will change then, but... it's hard for me to get excited about using a worse product today on the promise that it might get better in the future. And I guess it's accurate enough that a bunch of people keep telling me that they're saving time when they use it, so maybe I don't understand what I'm talking about. But I just don't see how people are reaching that conclusion unless they're either asking questions where they don't actually care about the accuracy or unless they're just rolling the dice and trusting that ChatGPT won't accidentally poison them when they ask what drug combinations they can take.
Just for clarification, do you mean the actual encyclopedia, or Wikipedia?
I think that encyclopedias are cool, but I also feel like the Internet was hopefully going to be a slightly better version of that and that it's a little frustrating to be going in the opposite direction. I'm not sure how to articulate that other than that encyclopedias are in some ways a compromise around the fact that we very often don't have good search, so we accept human beings trying to pre-aggregate data for us with the hope that they are better than Google is at aggregating that data.
And I think Wikipedia is valuable not so much because of the summaries, and more because it's obsessively curated and has (or attempts to have) a very specific, predictable set of rules that it (tries) to adhere to about sources and coverage. The text-portion of Wikipedia isn't really the part that I think is most impressive about it. If GPT-3 was being used to aggressively curate search results and remove low-quality content, then yeah, that would be potentially interesting (although I'm not sure how well it could handle that task).
----
> what can I make with kale, tomatoes, and mushrooms?
I sort of see it, it's one of the more... it's fine. It wouldn't be a strict downgrade over existing search, maybe it would save time in some situations. But if I'm being completely honest, what I want as an answer to that question isn't a paragraph of text explaining multiple recipes, it's a bullet-pointed list of recipes with links to the original sites they're listed on, so I can check to see if they're worth making.
Bing's results (to its credit) seem like they're sort of headed in that direction, which, nice. Yeah, I could use that. But a bullet-pointed list of recipes is also... search results. So how much time and effort have we put into reinventing a cleaner search interface where at best it solves the same problems that search already solved, and where we in the meanwhile haven't made any progress on the really hard fundamental problem of "how do we get a good list of recipes to display in the first place and what does 'good' mean in that context?"
----
To be less cynical, one way I could see this genuinely being useful would be completely behind the scenes in a non-user-facing role just as a way of applying "tags" to webpages or doing filtering. I would love to be able to search "is spinach poisonous for rats who are pets and just because I used the word 'poisonous' does not mean I want 2 pages of links about exterminators or getting rid of wild rats."
An LLM would be a great fit for that, because we've done such a garbage job of categorizing the web that it's very difficult to know which words to type to exclude "categories" of information from a search query. So maybe an LLM helps standardize that a bit? But then I want a normal page of search results. I don't want to have a conversation with the LLM, I want the LLM's role to be exclusively "I think this webpage should be additionally included/excluded in your query. I think this webpage is about extermination even if it doesn't use that specific word."
If Bing's service ends up being able to do that kind of thing well, then yeah, that's useful. I'm a little skeptical they can, because... gestures to the current quality of search results, but maybe their integration with GPT gives them way more capabilities and ratchets up their quality.
But similar to above, it feels a little bit like reinventing the wheel. I can refine GPT queries during a conversation, great. Could I have that feature for regular search? Why do I need to do it as part of a conversation? That seems like a good thing tied to a bad UX (although again, I might be atypical in thinking that natural language conversations are often bad UX).
And I do want to couch that by saying that maybe Bing will surprise me and it will have easy options to do that kind of thing. I'm just not currently seeing it presented in a way that looks useful.
Fully agree, but the problem is not really specific to AI. Most search results today, whether written by humans or machines, tend to be mindless, inaccurate, and unhelpful n-th hand summaries produced with very little effort.
And this doesn't seem unreasonable, given that you get everything for free or at a very low cost. When you pay for high quality books or periodicals, you get in return much better sourced information written by people who know a lot more about the subject they're writing on than the average journalist or AI language model.
Yes, occasionally one might find high quality contents in blogs, forums, wikis, or open-access periodicals, but far more are locked inside proprietary platforms or behind paywalls that do very little to actually compensate the authors.
Search engines and content platforms are supposed to make it easier to find what you want. But the reality is that it's a lose-lose situation for both the writer and the reader. The writer is forced to give up their contents at a very low price and overpay for ads, while the reader is left with low quality contents that aren't relevant to their needs. But neither can escape the monopoly, who alone profits at everyone else's expense.
Readit News
From what I know about Google, they are serious about least privilege type of stuff internally and employees dont get arbitrary unbound access to systems or data.
Just like the OP I have first hand info on how atrocious Microsoft’s internal privacy controls are.
The later versions of Windows are just ad space for Microsoft to advertise.
Azure is the worst of the 3 cloud providers. Horrible developer experience and documentation and reliability.
That said, I would still take Satya’s Microsoft over Balmers any day
I'd agree they are both evil in that they both seek to leverage your personal data against you for their own gain, but it sounds like Google does it while being more protective of what they've taken from you while MS also leaves you vulnerable in new ways through carelessness.
I'm not too worried about Bing though. That's nothing compared to Microsoft's access to your computer at the OS level. We know anything we do online will be seen by others, but being able to snoop on our personal files and log our keystrokes is a whole lot worse than knowing what we type into a search engine.
However improved/more popular Bing becomes, eventually Bing will end up in the same position Google is in now where they'll have to make their web search worse in order to constantly shove ads in your face instead of returning results that are useful, and slowly SEO spam will adapt to pollute Bing's results further. Google may not even mind if Bing starts gaining users. At this point Google can collect so much data from android devices that I doubt they need google search to peer into our lives like they used to anyway.
Really? I find Microsoft's documentation for Azure (in English) to be quite thorough and helpful. Their tools are well-designed and quite powerful.
Perhaps those of AWS and GCP are even more amazing, but I wouldn't call Azure "horrible" by any stretch of the imagination.
Microsoft is evil v1. Google and everyone post 2000 is evil v2.
They are just as evil as Microsoft but they've learned that amongst other things, you need to a) seem nicer and b) lobby politicians.
I'd argue that made them more evil.
Though it has to be said that Microsoft has learned, too, so at this point they're pretty much the same thing.
Don't forget security! Multiple serious and extremely trivial cross-tenant security exploits only in the past couple of years. AWS and GCP have had none ever, the much smaller OCI has had one.
Lets not use emotionally driven, ambiguous terms like 'evil' to describe these kinds of things. Google is not, for instance, conspiring to commit genocides, nor is Microsoft planning a coup in an Eastern European nation with the goal of colonizing them to use as indentured technical support.
Microsoft is the capitalist evil. Google is the technocratic AI evil. The first wants your money, as much of it as they can hoover, using dark patterns if useful for it. The second wants to optimize your life using _their_ metrics, without ever engaging with you (humanity is the problem - engaging with it never scales).
And I agree, Google does take security more seriously than most places.
IIRC there was some basic safeguards though, like a query must have had like 10 unique occurrences or something. I also have no idea if you could tie searches to people.
I was just looking at aggregate numbers. That was really cool. It was like Google Trends but with real numbers!
Why?
For one, I've had a Microsoft account for over 20 years now (anyone remember Hotmail?). It's long past due for me to be complaining, and Microsoft hasn't wronged me in that time anyway. By comparison, my oldest Google account only goes back just over 10 years, and horror stories abound even if I've been fortunate so far. I keep all my truly important correspondance and login tie-ins with Microsoft (read: my Hotmail).
For another, Microsoft nurtured over 30 years' worth of good will from me with Windows and Office; even though I hate many things about Windows from 8 and up, among other things, I will ultimately be a friend to Microsoft simply because they were a significant and positive part of my childhood and now my adult life.
These are companies, not family friends.
From memory, Microsoft used to be absolute bastards when it came to your data. They did everything they could to keep you gated in and without any control. Things like: refusing any export tools for your emails so you couldn't migrate to another provider; or, not allowing any non-web access to your email.
Then Google came along with GMail and revolutionised things: amazingly generous free data allowance, IMAP access, data download options, open standards, liberal approach to letting people set up various hacky app access, and so on. It was hugely successful for them, and forced MS to moderate their approach.
Google abandoned their Don't Be Evil strategy and sacrificed a lot of their principles (and, I would argue, a lot of their success with it), but at one point they had a positive impact and they've never been as evil as early-era Microsoft.
The monopolistic dominance is the issue. They have too much data and too much power, and leverage that to swallow other market categories. Today they have a deeply unhealthy dominance over the tech industry.
Google could do everything right and still be an existential problem. You say MS has data security issues, okay sure. I’d still rather have them or anyone at 25%+ of the market.
Google isn't forcing someone to use it. It's monopoly can disappear when something better it's there.
You're responsible for your data, not Google, not Microsoft.
In terms dev-related stuff, TensorFlow - literally something that's made today's ML possible, Kubernetes, BigTable, MapReduce, and yknow, that good old thing called Chrome..
I understand hating on Google. They're not that great, but none of the FAANG are "great".. They all have some terrible bits, some good bits. However, to say that Google has not provided anything except "spam stuffs" is very wrong.
Search alone has changed the entire way that humans consume information. I would say that today's internet landscape, for better or worse, would not exist without Google.
I wonder if the article authors were thinking of Xerox PARC's offices in Palo Alto's hills? Or the road named "Sand Hill Road" that used to have some venture capitalist offices? Other than those offices I can't think of any significant hill-based offices in Silicon Valley.
Well ... Not a big one, I grant you, but hill nevertheless :
https://www.google.com/maps/@37.4231911,-122.084872,3a,75y,1...
Although Google, Maps, Youtube are of daily use they are monetized exclusively by advertising which is annoying and hated by many. It has been many years since Google has launched an innovative new product.
I don't think ChatGPT will gain daily traction after this hype. Anyway we could say that MSFT and AMZN have demostrated more power to innovate with different business models (not only adv) and products.
The GOOG stock has a PE (Price/earning) of 23, while Coca cola 26. So the stock market expect higher growth from CocaCola than Google. Quite surprising.
- GOOGLE PE (23): https://www.macrotrends.net/stocks/charts/GOOGL/alphabet/pe-...
- COCA COLA PE (26): https://www.macrotrends.net/stocks/charts/KO/cocacola/pe-rat...
> Employees were encouraged to spend up to twenty-percent of their paid work time pursuing personal projects. The objective of the program was to inspire innovation in participating employees and ultimately increase company potential. For Google's part, Gmail and AdSense both arose out of side projects. In 2013, Google discontinued 20 percent time.
What has Google released since 2013[0]? Stadia?
Then you tack on that Google Search as of today is basically unusable/broken with people needing to append site:, double-quoting everything, or enabling verbatim search to get high quality results. Heck, on top of that, I also have a uBlock configuration that removes 280 domains from Google Results because they're all auto-generated spam of Stackoverflow answers.
There is something deeply wrong with the culture inside Google.
[0] https://en.wikipedia.org/wiki/Google#2012_onward
- GCP Suite (a full integrated and consistent suit, answering so many things)
- Google Workspace (trust me when you get to know everything that is possible you will be amazed)
- Tensorflow and other Machine Learning applications
Can you describe exactly what you mean? If there is a way to hide specific domains from showing in search results, it would be an incredible boon for me.
The press loves a David and Goliath story; the young disruptor versus the stodgy disrupted.[1] At that time, Dropbox was David, coming out of nowhere in a hurry and had a product that seemingly should be so obvious for Google to launch but hadn't - making Google seem antiquated and slow.
Fast forward a couple of years (aka a decade) and Dropbox is still going ok but Google Drive is by far more ubiquitous.
I guess only time will tell.
[1] Dropbox Versus The World https://www.fastcompany.com/3042436/dropbox-versus-the-world
I love chatGPT but it clearly has a ton of issues. To declare Google dead that is point is completely absurd.
I can't think of a single transformative technology that actually has a first mover advantage. The first, second and third mover seem to make all the mistakes and setup the forth mover to learn from those mistakes for free and create a monopoly.
Counting out a company that owns Deepmind is just not rational.
Here’s a tour of the recreation https://maps.app.goo.gl/syHHp9GWmaGA5Woz7
At the same time, a lot of these platforms ran effectively unopposed for many years, and now competition in advertising is spreading quite rapidly
Deleted Comment
I find YouTube quite easy to use, esp the shortcut keys.
Recommendation system is another story.
Youtube Music is really .. strange.
It looks wrong/outdated everywhere.
Generated playlists (recomendations) are always (for days/weeks) the same. Not a bit randomization. It feels like they were approved by party commitete for years.
I don't understand why they decided social network in music service.
Ah hell. CUE-lists are still too novice technology for them.
I still think that last.fm 15 years ago was the best music service.
Even Google winning the AI wars leaves them worse off as the operating profit potential of querying an AI vs their search index has to be far lower.
As to usage, once you start to integrate ChatGPT into your workflows it can meaningfully benefit vs traditional search. Ive been able to find information on specific programming language concepts, with generated examples, far faster than searching.
I can ask it about GameDev concepts and ask for bulleted lists or higher/lower level of detail in the answers. Information is presented in a much easier to consume manner
That being said, most stocks that are considered defensive are quite overvalued on a fundamental basis. I would consider a PE of 26 for Coke quite undesirable, though there’s much worse
In some ways, it might even be good. It's suffering from a kind of advertising resource curse nowadays, and being forced and able to diversify its economy would be best in the long term.
I don’t know how much. Money it makes, but there is paid API access to maps.
Since I can’t paste screenshots from my terminal, you can see the sector P/Es on page 6:
https://www.spglobal.com/spdji/en/documents/performance-repo...
I can say the exact same thing about countless other large companies: Facebook, Intel, Microsoft, Boeing, AMD, AT&T, I could go on and on. At some point, launching innovative new products isn't really that important for a company, when many people on the services that company provides.
Disrupt yourself before someone else does.
I'm not "blaming" them, it's the cycle of life. Gravity eventually wins.
This is also a key reason for all the layoffs: Growing, innovating companies cannot ONBOARD staff fast enough, because their value is decided by how big they will be in the future and how fast they can get there. People are a profit centre. But mature companies' (in mature industries) only source of increased value is in efficiency, getting the same revenues with less costs. So people are a cost centre and they succeed by showing they're efficient (aka laying off workers).
Stadia launched in 2019.
Also, it's already dead.
Stadia might have been well executed from a technical standpoint but AAA game streaming wasn't exactly a new and innovative idea in 2019.
Lol.
The article aims to offer a preliminary analysis of whether Microsoft can become a better parasite: Grabbing the content people generate online, paying nothing, and using it deftly to serve advertisement on the basis of private behavioral traits that are gleaned by prying open and subverting the use of all pieces of IT people use.
It is not too difficult by now to imagine alternate tech universes (Philip K Dick style) that have nothing to do with this nightmare, where more or less the same technologies empower individuals and companies and organizations rather than squeezing them dry. The combination of oligopoly status, moral laxity and political dysfunction means we are simply sitting around like sheep discussing whether a new butcher is about to get sharper knives.
The web was hard to navigate and you relied on webpages "befriending" each other and helping you navigate to similar pages. That was real navigation and it had terrible recall. Once google became the Internet's frontend everything was one or two hops away. It was awesome, and I think that it has evolved in a way that saves time with the snippets and knowledge graph panel at the expense of taking away clicks from sites that needed to serve their ads. Although today things are getting worse with walled gardens and SEO.
A problem that the internet has not solved yet is how to keep webpages up while there's increasingly more content. Most sites were maintained at a loss by people, or leeching resources at a university or company. The closest "people" have got to fix this IMO is through P2P, but it got too focused on piracy and got a bad reputation with malware too. And I think we never had the infrastructure to serve and share webpages we visited. Sharing today is still a mess and we still depend on centralized distribution + caching (and that's dying with https, only big companies who are better off giving caching servers away to ISPs can do this).
Ads came to "save" the internet from this problem. I really hate them and I wish we had a better model, but they seem a necessary evil as people just got used to believe that all digital things are free when they obviously cost something, and that still shows today in the apps stores.
Advertising worked just fine for more than a century, ever since the invention of consumerist society. (How much of that is really necessary in a society that is pushing headlong into unsustainability is another, bigger story.)
The disaster we have now in our hands is of a more immediate and crass nature.
Think about it: there are algorithmic breakthroughs that can help everybody on the planet raise their level of information retrieval and the only discussion is which of the former or the current consumer tech oligopolies will use it best to push ads and how to double down on a bad design.
>the web was hard to navigate and you relied on webpages "befriending" each other and helping you navigate to similar pages. That was real navigation and it had terrible recall.
This is exactly why internet was decentralized. A network of social connections between small online spaces people congregated in was essential to find what you wanted. It was somewhat of a replication of the old world - where in order to find an answer, you first had to get into relevant circles (eg. a local club, university, etc). You didn't just get an answer to your question, you inevitably had to sift through a lot of other content and learn about new places.
Google cut through all of that. Ideally, you could just get the answer you wanted, and not even see the rest of the site. The first order effect was that everyone's life get better. The second order effect is that it killed the old decentralized web, because random discovery nearly died down.
LLM are only the next step on this, but I don't think they change much. I think it's mostly going to damage reddit. A lot of reddit questions are an attempt to find an answer to something that google is too dumb to find - but llms may be able to.
The defense to this - are chatrooms, like discord. Sometimes I see people complaining that so much information is now 'hidden' on discord. But that's the exact point. Making information hard to get means people are forced to interact with each other. This creates incentives to contribute.
That's the future. Chatrooms may be replaced by voice based vr, hard to say. We already passed the peak of public information dissemination, and are going back to old style "ask at the university", just more decentralized and online.
Don't tell me you forgot Yahoo, Excite, Altavista, Lycos, and all those other search engines existed! Lycos also had hotbot and tripod.com, back when homepages were your internet presence instead of social media.
Maybe Google had results that were a bit more relevant and won that competition, but they didn't Change Everything.
In this model, are end users who use it to search for information parasites as well? After all, they are consuming content that other people generate online, and usually pay nothing for it. And they love to look for — and find — what they are interested in. If Google didn't satisfy that need well, people wouldn't have preferred it, and it wouldn't have grown.
Users pay with their data. That is the whole point of the business model and it is obviously lucrative enough so that the entire tech infrastructure (devices, OS, browsers etc) can be repurposed to be a user data collection channel. If you want to find a real accomplice that is essential for the model to work, it is not the users, it is the advertisers. It takes two to tango in the adtech market.
I don't dispute that search (in its various incarnations) is an essential service in a digitally interconnected world. There are countless ways to pay for it (as a digital public good, as a user subscription etc) that are fundamentally better than what we have. It is also obviously true that some decades ago Google innovated technically. A lot happened since and it wasn't positive. Normalizing it simply prolongs the agony.
Search should be democratized by governments. It is central and should be for everybody. No control.
These LLMs will be the end. They will capture enough of the user attention that the few remaining ad supported sites will be wiped out. “Content”, that low quality grist every site is filled with to capture users for a few seconds before they click on will become worthless, no different than the low quality crap the LLMs spew out.
Finally, we will stop trying to keep the bloated corpse of the web alive and will move on. It will be the death of a dream. At least we can all let go and focus our energies on a new dream. Whatever that dream is, I hope high quality, human created art, knowledge, opinion and creativity is at the center of it.
Deleted Comment
Speaking of that, I might as well provide some supprting evidence from a different part of the tech industry. Most people don't know that right now Blackrock and Coinbase are attempting a coup on the US dollar by seeking access to central bank liability for their stable coin? The Coinbase CEO even said it out loud that this year his stable coin will become the de facto central bank digital currency. I suspect that they have indirectly bribed someone at the Fed (how else does that work? esp since the Fed chairmen has come out strongly against stablecoins) and are basically trying to get the ability to get Fed loans at the Fed rate and loan out at the commercial bank rate without being a bank and by using a coin which they can mint themselves, so all the profits from interest create more reserves and the cycle keeps going. It is literally insane. The threat from Google and Microsoft to society pales in comparison. I may not have my understanding 100% correct but all banks are up in arms about it and the average person on HN has no clue this is happening. Just Google Blackrock, USDC and the Fed. The banks are calling it a backdoor CBDC.
Yes, Blackrock and Coinbase have partnered up to provide crypto trading access to Blackrock clients but how are they attempting a coup on the US dollar?
The correct interface for ChatGPT + search is just...ChatGPT. But it can also show you a list of web search results, when it's appropriate.
A super-clean interface, that always shows you exactly what you want.
That would be a killer feature and represent a real threat to Google.
https://blogs.microsoft.com/blog/2023/02/07/reinventing-sear...
At least one person is working on this:
https://twitter.com/hsu_steve/status/1623388682454732801
While something like this would of course be incredible, even just current ChatGPT + the ability to web search for you would be quite amazing.
If MS isn't ready to completely replace Bing, then they should've made it an entirely separate page: bchat.com or something.
Rare: 120-125°F (49-52°C) Medium-Rare: 130-135°F (54-57°C) Medium: 140-145°F (60-63°C) Medium-Well: 150-155°F (66-68°C) Well-Done: 160°F (71°C) and above
It's important to note that cooking steaks at low temperatures for extended periods of time can result in a more tender and flavorful finished product, compared to cooking steaks at high temperatures for a shorter amount of time.
Looks great to me.
I guess they opted for the search results as the default with the chat on the side as the link list is the current interface.
Deleted Comment
That's the pipedream, but in reality isn't the case yet even using just ChatGPT.
Bing's implementation is trying to balance raw ChatGPT outputs, with summarised cited responses in search.
I do think they need to speed up the animation for generating the summarised or ChatGPT result though...
Google, has all three. The real question is can they not shoot themselves in the foot while doing it.
Also SEO will always target the market leader.
As far as Bing's results: I keep thinking Google is trashy. Then I use Bing and I remember just how good Google is. That's my personal opinion. I am NOT going to claim they are SEO, and spam immune, far from. But they seem to do better than Bing in getting what I mean right.
ChatGPT may help MS, but is it a bandaid? Is it just a good PR bump? Who knows... Displacing Google is a large task.. and not one I'm sure I want MS to win. But I look forward to them trying, if only to make Google do better.
The other searches are worse. I too have tried. But in general I'm frustrated with Google now more than I've ever been.
I'm talking about lists like these: https://github.com/quenhus/uBlock-Origin-dev-filter/blob/mai...
It wouldn't take a small team more than a few weeks to get through these lists -- most of the work has already been done.
Alternatively, they could just let us add these lists to a first-party interface so we don't have to use plugins to achieve the same results.
I use a lot Ecosia (which gets results from Bing) due to them allegedly planting trees and being much more heavily into privacy.
My experience with Ecosia, thus Bing, is that for the overwhelming majority of searches (90% plus) the quality is comparable.
When you get into more niche searches Bing offers better results 1 out of 5 times and worse results 4 out of 5, with a worrying amount of those 4 being almost complete misses.
I still need to get onto Google occasionally when I don't find what I was looking for on Ecosia.
MS has them all three too.
And since, as you say, the algorithm part is not the most difficult part, others may enter the playing field soon (we already have DDG and Kagi, for example, with Kagi now experimenting with LLMs too).
MS has the money. Data is by far and away the hardest part, and DDG's own privacy policies etc will hurt it here, alas.
Personally, I think it is Google's race to win. But they have to DO it. If I start getting better results off DDG or Bing, I'll switch. I'm fickle, and own no loyality to Google.
I just want my StackExchange answers faster, before Google Coder, and Co-Pilot replace me. /s
The problem with search is not that our answers aren't summarized well, it's that the quality of information returned for those searches is getting increasingly worse, and we are getting increasingly worse at categorizing or filtering that information in any useful way. And LLMs pulling information in and summarizing it for me is... not helpful? It's summarizing the same garbage, except now sometimes it also summarizes it wrong.
But it's not even an issue with the quality (although the quality of information from LLMs is also pretty over-hyped I think). Conceptually, I don't know that this is a product that I would ever want. I can't think of any time where I've sat down to do a search on Google or DuckDuckgo and thought, "You know what I want? I want these results presented to me in a less structured format using natural language and with less granular knowledge about where each specific statement is coming from."
At least Bing seems to be trying to do inline citations in some of its answers, which is a step up over Google's AI announcement, I guess?
Maybe I'm just in the minority on that. Users seem to like this a lot. But my ideal version of the Internet is one that decreases the number of abstractions and layers and summaries between myself and primary data rather than increasing them. My ideal Internet is a tool that makes it easier for me to actually find things, not a tool that increases the layers between me and the raw source/information that I'm looking for. I already have enough trouble needing to double-check news summaries of debates, events, and research. Getting another summary of the summaries doesn't seem helpful to me?
I can think of some ways where I might use an LLM in search, even really exciting ways where maybe it could help with categorization or grouping, but it doesn't seem like Google/Bing are interested in pursuing any of that. I look at both the Bing and Google announcements and just think, "why are you making it worse?" But who knows, maybe the actual products will sell me on the concept more.
The first is a little more mundane: LLM embeddings. OpenAI currently offers an API that turns sentences into coordinates for a point in some 1536-dimensional conceptual space such that two points are close together if they are conceptually close together. This is insanely powerful. For example, you can generate captions for a bunch of images and store the embeddings for them. Then, you can look for a "picture of a rabbit eating a carrot" by turning that phrase into a 1536-dimensional point and looking for the nearest points around it. Basically, it blows open search technology for everyone. You no longer have to deal with synonyms, idiomatic phrases that mean similar things, misspellings etc - the problems you'd run into when trying to implement simple text search using traditional techniques. It all gets simplified to generating coordinates in some hyperspace and looking for nearest neighbors. This is a total game changer.
The second direction is ChatGPT. Sure, if you want to read a detailed analysis of the demographic situation in China, you'd prefer an article written by an expert. You would still use a search engine, pick a search result and do things the way you do them today. However, there's an entire collection of things that can be answered directly by ChatGPT. For example "how many mins should I hard boil an egg" or "Can I take NyQuil when I'm stoned" or anything else where you really just want a single sentence answer. Today, you launch a browser, search for what you want, skip past the first 10 advertisements, look for a site that seems reasonably reputable, click through all the GDPR warnings, scroll past the banner ads and the SEO optimizing bullshit text to find that one sentence that you wanted all along. Or, you could ask ChatGPT and get an answer instantly. (assuming chatGPT is good enough eventually).
It's hard to predict which of these two technologies will disrupt the current status quo in search. Neither might. But we haven't ever been this close to a level playing field in search since the 1990s. The excitement is hard to resist.
You know Google has been doing this for years now?
> However, there's an entire collection of things that can be answered directly by ChatGPT. For example "how many mins should I hard boil an egg" or "Can I take NyQuil when I'm stoned" or anything else where you really just want a single sentence answer.
Google has been doing this for years via search cards, which are AI generated summaries of website information.
> there's an entire collection of things that can be answered directly by ChatGPT. [...] assuming chatGPT is good enough eventually
I am a lot less impressed with this. And I know I'm an outlier and plenty of people are shocked at how good GPT is at this kind of problem, so I am constantly second-guessing myself and thinking to myself, "are we using the same product?" Because I think ChatGPT produces really bad quality information. It's cool, it's wildly impressive, it's a massive achievement and an incredible milestone for AI, but 'cool' is different from 'useful.'
Leaving aside the problem that answering simple questions is a very small subset of what search is used for, and isn't on its own probably a big enough category of questions to make me change search engines, the bigger problem is that the current state of ChatGPT seems to be wildly inconsistent about what it knows and what it doesn't know and I don't have a way to pre-predict what categories of information it's safe to ask about. And the only way for me to verify the answers it gives me are to... double check its work with a real search.
I would not advise anyone to ask ChatGPT for advice about what drugs are safe to take while high, that seems profoundly unwise to me.
So it's a bit like Instant Answers. Google has been trying to auto-answer questions for ages, and in practice the only time it's ever been useful for me is when it's extremely predictable and when I know that a category of question will only ever have its answer pulled from one site and where I know what the format of that answer will be.
Unpredictability is generally a quality that I try to avoid any time that I am using a computer. One of the primary strengths of a computer to me is specificity and predictability. And so the bar here is really high. The question I ask myself is, "would I want to replace a search engine with a human assistant?" And I think the answer is no, I feel like that would be missing the point of what a search engine is. And ChatGPT gives worse answers than a human assistant would, and its sources/knowledge is just as unpredictable as a human's would be if not worse. So, I also don't want to replace my search engine with ChatGPT.
It could get more accurate in the future, and if it does then maybe my opinion will change then, but... it's hard for me to get excited about using a worse product today on the promise that it might get better in the future. And I guess it's accurate enough that a bunch of people keep telling me that they're saving time when they use it, so maybe I don't understand what I'm talking about. But I just don't see how people are reaching that conclusion unless they're either asking questions where they don't actually care about the accuracy or unless they're just rolling the dice and trusting that ChatGPT won't accidentally poison them when they ask what drug combinations they can take.
I do see LLMs as potentially more useful for “fanciful” queries, like “what can I make with kale, tomatoes, and mushrooms?”
I think that encyclopedias are cool, but I also feel like the Internet was hopefully going to be a slightly better version of that and that it's a little frustrating to be going in the opposite direction. I'm not sure how to articulate that other than that encyclopedias are in some ways a compromise around the fact that we very often don't have good search, so we accept human beings trying to pre-aggregate data for us with the hope that they are better than Google is at aggregating that data.
And I think Wikipedia is valuable not so much because of the summaries, and more because it's obsessively curated and has (or attempts to have) a very specific, predictable set of rules that it (tries) to adhere to about sources and coverage. The text-portion of Wikipedia isn't really the part that I think is most impressive about it. If GPT-3 was being used to aggressively curate search results and remove low-quality content, then yeah, that would be potentially interesting (although I'm not sure how well it could handle that task).
----
> what can I make with kale, tomatoes, and mushrooms?
I sort of see it, it's one of the more... it's fine. It wouldn't be a strict downgrade over existing search, maybe it would save time in some situations. But if I'm being completely honest, what I want as an answer to that question isn't a paragraph of text explaining multiple recipes, it's a bullet-pointed list of recipes with links to the original sites they're listed on, so I can check to see if they're worth making.
Bing's results (to its credit) seem like they're sort of headed in that direction, which, nice. Yeah, I could use that. But a bullet-pointed list of recipes is also... search results. So how much time and effort have we put into reinventing a cleaner search interface where at best it solves the same problems that search already solved, and where we in the meanwhile haven't made any progress on the really hard fundamental problem of "how do we get a good list of recipes to display in the first place and what does 'good' mean in that context?"
----
To be less cynical, one way I could see this genuinely being useful would be completely behind the scenes in a non-user-facing role just as a way of applying "tags" to webpages or doing filtering. I would love to be able to search "is spinach poisonous for rats who are pets and just because I used the word 'poisonous' does not mean I want 2 pages of links about exterminators or getting rid of wild rats."
An LLM would be a great fit for that, because we've done such a garbage job of categorizing the web that it's very difficult to know which words to type to exclude "categories" of information from a search query. So maybe an LLM helps standardize that a bit? But then I want a normal page of search results. I don't want to have a conversation with the LLM, I want the LLM's role to be exclusively "I think this webpage should be additionally included/excluded in your query. I think this webpage is about extermination even if it doesn't use that specific word."
If Bing's service ends up being able to do that kind of thing well, then yeah, that's useful. I'm a little skeptical they can, because... gestures to the current quality of search results, but maybe their integration with GPT gives them way more capabilities and ratchets up their quality.
But similar to above, it feels a little bit like reinventing the wheel. I can refine GPT queries during a conversation, great. Could I have that feature for regular search? Why do I need to do it as part of a conversation? That seems like a good thing tied to a bad UX (although again, I might be atypical in thinking that natural language conversations are often bad UX).
And I do want to couch that by saying that maybe Bing will surprise me and it will have easy options to do that kind of thing. I'm just not currently seeing it presented in a way that looks useful.
And this doesn't seem unreasonable, given that you get everything for free or at a very low cost. When you pay for high quality books or periodicals, you get in return much better sourced information written by people who know a lot more about the subject they're writing on than the average journalist or AI language model.
Yes, occasionally one might find high quality contents in blogs, forums, wikis, or open-access periodicals, but far more are locked inside proprietary platforms or behind paywalls that do very little to actually compensate the authors.
Search engines and content platforms are supposed to make it easier to find what you want. But the reality is that it's a lose-lose situation for both the writer and the reader. The writer is forced to give up their contents at a very low price and overpay for ads, while the reader is left with low quality contents that aren't relevant to their needs. But neither can escape the monopoly, who alone profits at everyone else's expense.