Agreed a few months ago I had multiple daily conversations with ChatGPT. Anymore I might open it once in a day if that and usually just for some "generate a short presidential speech about why (this song) is cool" or something.
Stable Diffusion is a different story I'm constantly generating something but it's more about playing with the tech than actually making a meaningful product for me. I just enjoy seeing what cool stuff I can make it do.
I think Google has realized Open AI's ChatGPT is its own product and not a direct competitor to Search, although there is some overlap, which is why there is Bard as a direct competitive product to ChatGPT.
Search is its own beast, and I suspect they have the data to prove users want something different from a Google search.
> OpenAI has a tremendous lead in terms of quality of the product and adoption
> On top of that, thousands of new models and projects are coming out almost every day that are starting to rival the quality of GPT as well as coming up with amazing, completely new things
I've read ChatGPT growth has already flattened [0]. Sure thousands of new models/projects are coming out, but until the general public starts to use them I don't see this as much of a threat to Google.
On it's own, ChatGPT makes up answers. Bard does as well, but it's worse. Google's integrated featured snippets are notorious for wrong answers. When Google says they are adding more AI to their search experience, I'm not optimistic.
I've been using perplexity.ai for the majority of my searches in the past month. It isn't ChatGPT but it uses OpenAI's GPT APIs. It sources answers from search results and cites them. Even if the answer may be wrong, it came from other websites and it didn't just hallucinate them based on memorized data.
It is much faster to find answers this way. It certainly meets the magnitude test for speed and quality -- I get answers at least 10x quicker. Additionally, because it is reading through multiple results you can see when different sources are providing conflicting information.
Google has the capability to do all of this. The unknown is what does it do to all of the ad revenue being driven by the users who have had no idea they've been clicking on ads to the tune of $200+ billion a year?
Does that mean there will be too much internal resistance to revenue cannibalization when other platforms have a better search experience?
> It sources answers from search results and cites them. Even if the answer may be wrong, it came from other websites and it didn't just hallucinate them based on memorized data.
You can get this same experience, with GPT by using the Edge browser/app
It will give you an answer and at the bottom the sources with links so you can check. Really helpful
I'm in the middle of listening to this book and they talk about "Staff+" roles some inside of it. The books describes the blurry lines in there, but it does talk about that next level in general.
Right, so OP wants to treat Bard as an API. I don't dispute the thought, though it does run counter to conversational models. I'm slightly anxious about all the rogue conversational text parsing to use LLM's as an API.
The only reason they’re having a problem is due to the way they’re requesting this. OpenAI’s GPT can follow schemas for responses and issue commands or API requests if you tell it how. I’m sure Bard can do the same with the right request structure.
Yes, white culture. Black people don't find that kind of stuff funny or cute. Middle Easterners don't. Most Europeans don't. Africans don't. South Americans don't. Asians don't. Some out-of-touch, white Americans do. That's the target audience, I suppose. It's snarky and hollow.
All white people aren't upper middle class coastal people. There are also additionally upper middle class coastal people that are black and middle eastern and like upbeat tech videos. Begone with your racism.
it's not necessarily the target audience, but it's who's making these videos generally — mainstream american culture has always had a major yuppie component as a primary feature, google has embodied this for quite some time
Cloud turning a profit is a good thing, and makes sense. But cloud is also a super competitive market - not so for Search or Ads. Those two feel like the places Google can rely on money coming from - revenue falling for those is worrying!
I would question just how competitive cloud is. Companies with banks worth of cash and an existing army of technical knowhow are few and far between. The only companies that could enter the market right now are facebook and Apple.
Ad spend globally may be down; can a comparison be made to competitors' ad revenue to come up with a baseline there?
Youtube was a large contributor to growth in 2021-2022, and that engine appears to have slowed. (Not just the "Youtube ads" line that is separated out.)
Most likely the market is reacting to these results not being any worse. Cloud gains were already priced into the stock, in my opinion.
"Cloud" for the purposes of financial reporting also includes the consumer cloud services workspace/gsuite/gdocs or whatever they're calling it now, right? not just the stuff at cloud.google.com
From the report: "Google Cloud includes infrastructure and platform services, collaboration tools, and other services for enterprise customers. Google Cloud generates revenues from fees received for Google Cloud Platform services, Google Workspace communication and collaboration tools, and other enterprise services."
> - unallocated corporate losses are 3B higher YoY
$2.5 billion of that is charges related to the layoffs.
> - all their expenses are way up (cost of revenue 1B up, R&D expenses 2.3 B up, sales and marketing costs ~0.7 B up)
Likewise half of this (i.e. you're double counting). See the table on page 3.
> - ad revenue is down (that's 80-90% of all of their revenue)
78% is not 80-90%. It's impressive that you managed to quote a ridiculously large range, and still get it wrong, in a discussiong about the earnings report.
The ad revenue from Google properties, not from the display ad network, is up.
> I couldn't care less to calculate it this time. It fluctuates around the same number YoY
Uh-huh. It was 77% in Q4, 79% in Q3, 80% in Q2. It has not been 90% for a decade.
> And that somehow makes it less of an ad revenue? Or different? Or something?
So you weren't making any kind of point when saying that ad revenue was down? If that's the case, I'm happy to also pretend that I'm also a member of the non sequitur club.
And since there are no details what is included on Cloud, number of new customers, example of large accounts that moved to Cloud, or info how much is Workspaces versus Enterprise Cloud customers...the moment those details leak, or if this is shown to be financial reporting engineering, the market will turn very sower, very quickly.
The amount of fatigue I get having to determine if what they tell me are fact is just too much.
I'm sure someone will tell me my experience should be similar with generic web search, but at least I'm in control of what websites to read through to determine sources.
However, I'll agree with most that state it is helpful for creative purposes, or perhaps with coding.
I've found they serve almost exactly the opposite purpose as search engines. When I want reliable info and don't need hand-holding: search. When I have no idea what to search, or want a quick intro to something: ChatGPT. Together, they are very powerful complementary tools.
Your experience absolutely shouldn't be similar to generic web search. The idea that they are an effective replacement for that is one of the most widespread misunderstandings.
They're good at SO MUCH OTHER STUFF. The challenge is figuring out what that other stuff is.
> The challenge is figuring out what that other stuff is.
Unfortunately, the major problem is something you pointed out in your blog post:
> We must resist the temptation to anthropomorphize them.
The reality is that, we in meatspace simply cannot help but anthropomorphize them.
These language models regularly pass the Turing Test (admittedly for low bars).
They are surprisingly good at bypassing the Uncanny Valley to hit the sweet spot of persuading without legitimate justification, simply because they are so convincing in formulating sentences in a manner that a confident human would.
Yes, these tools have legitimate use cases (as you outlined in your blog).
But the vast majority of use cases will be those of confidante, of discourse partner, of golem brought to life without understanding what exactly has been brought to life.
I find it very useful for doing zero shot and few shot classifications of natural language input.
The "use it as a chat companion" is an interesting technology demo that demonstrates some emergent processes that make me wish I was back in college on the philosophy / linguistics / computer science intersection (though I suspect the hype would make grad school there rather unpleasant).
I don't think we've figured out stuff it's useful for, we've just created tech-demos that are much more digestible.
For blockchain/crypto companies their tech demos have required you having a wallet, downloading an app to interact with the chain, or just having lackluster visuals for the users involved in the tech-demo.
On the other hand, LLMs can be interfaced via strings in APIs, so it's braindead to spin up a text-interface for those APIs and no wallet setup or learning about new chains, the English that works on one model will work on another and produce results that are better than most cryptocurrency/blockchain tech-demos.
Notice that none of this relies on us having "figured out all kinds of stuff that this is useful for". We've made cool looking tech demos that make it easy for anyone to generate content.
Much like blockchains I feel it's the underlying technology that's actually useful(distributed PKI for blockchains and deep learning networks for GPT), and GPT itself is only 'useful' insofar as it's an easy-to-interface with implementation of a much more powerful idea.
I mean, the usual argument about why blockchains aren't useful implies they have to be useful for every person in all situations and that tradeoffs are unacceptable, so if there is some marginal extra cost or complexity then no matter how many benefits I might claim to be getting from using blockchain technology every single day as a replacement for random banking institutions I'd previously been having to deal with for decades that I'm somehow just wrong and there are no actual use cases...
..and that's the same deal for GPT as far as I can tell: you might think you are getting value out of it, but people such as maybe-literally-me are going to whine that the error rate is high and that people are not paying enough attention to how they are using it and that at the end of the day it is probably worse for you than learning how to do things yourself and that the whole thing is overrated because many of the things people try to use it for can be done by a person and maybe we should regulate it or even ban it because all of this misuse and misunderstanding of it are dangerous to the status quo and might be the downfall of western civilization as we know it.
To be clear: I'm using it (ChatGPT) occasionally for some stuff, but it hasn't replaced Google for me anymore than crypto has fully replaced banks... and yet the fact that I am using either technology as often as I am on a daily basis would probably have been surprising to someone 10-15 years ago. And yet, in practice, most of the stuff people are excited about in both fields is, in fact, a tech demo more than a truly useful product concept, and one that only is exciting momentarily until you get bored.
I think you’ve got some combination of a utopia fallacy and a straw man going on here.
I just want to contrast two things. First, blockchain had a lot of hype around utility that never materialized. It is really quite a minority that ever used it for anything besides buying it on a platform and hoping it would go up. The big adoption was always about to happen.
Second, ChatGPT is totally different from this. Its usage is not future tense. It is present tense and past tense. I can’t get across how different “someone will use this tomorrow” is from “someone used this yesterday”.
People are wildly excited about the future and things that haven’t been built. This does not change the fact that millions of people are using this every day to solve their problems. Saying “we haven’t figured out stuff it’s useful for” is just wrong.
Lately I feel like I’m at a park with people who are saying there probably isn’t going to be any wind today while I’m already flying a kite.
With Google search going steadily downhill, I find it really tough to verify anything that ChatGPT authoritatively states is true
Everyone on here is so enthusiastic about AI gobbling up the entire software landscape, I would just like a search engine that has any chance of telling me if something is factual
I've had your same experience. I've found them mostly to be an error-prone search engine, with somehow less accountability than the open internet, because it hides its sources.
At least with Stack Exchange answers, we have who wrote it, what responses there were, what the upvoting behavior around it was. And for the most part, I've found ChatGPT will transcribe often times wrong answers very poorly.
One small example, I asked it to solve the heat equation (i useded the mathematical definition, and not "the heat equation") with dirac initial conditions on an infinite domain. It did a good job of recognizing which stack exchange answer to plagiarize, but did so incorrectly, and after a mostly correct derivation, declared the answer was "zero everywhere."
It's kind of interesting that our science fiction projected traditional computing's strengths, math and logic, into the AI future with overly logical and mechanical AI characters. But our first creation of fully communicative AI has elementary school strength in these areas while it's probably better than the average adult at writing poetry or an inspiring speech.
I was mostly commenting on how it just plagiarized a correct answer off of Stack Exchange, except it took an incorrect hard right turn at the end to make up a solution.
This was me just testing it. I was aware of the particular SE answer ahead of time, and it followed the whole thing close enough that I had assumed it had internally mapped to it. But I suppose it didn't have to be that way.
It's like dealing with electricity (or maybe the internet). Early skeptics believe it is a curiosity with little application. People see how it can jump all over the place and create disasters that they can't imagine having engineered systems to finely control its behavior and create reliable complex functions and become the bedrock for computing.
I think there is also an aspect of willful disregard. This technology may change a lot, and it may be easier to dismiss that idea rather than process it.
Do you think there might be the opposite going on ? Wanting to believe something that isn’t there because you won’t have to do as much work, feel smarter etc ? Because it’s really hard not to anthropomorphize it ?
Gloss over all the incredible dangers we might be exposing our world too just because it’s “fun to play with” and see what AutoGPT can do to the Internet ?
I use it for thing's that don't really matter if they're exactly correct. For example, coming up with a travel itinerary for a country I have never visited. Rewriting a work email with better English. Summarizing a news article. There are lot of things that don't require ultimate precision. I feel like people expect these models to do something they aren't really designed for - and the mismatch in expectations causes people to be let down. They are just tools - not "mildly conscious beings" like OpenAI founders wants you to believe.
I asked it for references about Hafez Shirazi’s abandoned journey to India and it suggested a very specific Encyclopaedia Iranica entry which seemed perfect, and of course did not exist.
Ultimately, it's just a tool, so if the tool needs you to hold it this way and twist, you hold it and twist. And this seems to do the trick. Since it does answer with references for other situations, we needn't concern ourselves with the details.
The technical report[1] makes that claim at least:
>GPT-4 significantly reduces hallucinations relative to previous GPT-3.5 models (which have them-
selves been improving with continued iteration). GPT-4 scores 19 percentage points higher than our
latest GPT-3.5 on our internal, adversarially-designed factuality evaluations
It's much, much better than ChatGPT 3.5... in particular, if I ask it for biographical information about non-celebrity but internet-famous people I know 3.5 tends to make up all sorts of details while 4 is almost entirely correct.
It still makes things up though, just in less obvious ways. So the trap is very much still there for people to fall into - if anything it's riskier, because the fact it lies less means people are more likely to assume that it doesn't ever.
It still can't explain standard CS algorithms most of the time. I've just tried asking it to explain deleting a non-root node from a max heap with examples. And both attempts were either plain wrong (random nodes disappearing) or poor (deleting a leaf node which is not very illustrative).
Edit: I then asked who a certain deceased person _is_ and it gave me a completely wrong answer about a different person who's still alive and happens to share the last name. Both people have multiple website, books, publications and Wikipedia entries older than 2021 (which seems to be the cut-off).
Edit 2: Looks like I'm still on 3.5, so disregard the above.
The short answer is it can't. It's arguable whether anyone can - for a human being, determining if text is "factual" can be incredibly difficult.
A better answer: if a fact is present many, many times in training data - "Paris is the capital of France" for example, it's much more likely to be true.
Also influential: RLHF - Reinforcement Learning from Human Feedback. This is the process by which human labellers rate answers from LLMs - if they consistently rate up "facts" the models have a better chance of outputting factual information, at least if they can relate it to the up-voted responses somehow.
> The short answer is it can't. It's arguable whether anyone can - for a human being, determining if text is "factual" can be incredibly difficult.
Yet, most adults I deal with don't make false things up out of whole cloth as much as ChatGPT does, and it really does not seem like it is that difficult for them. Children do this quite often though, and some adults do, but most don't.
> A better answer: if a fact is present many, many times in training data - "Paris is the capital of France" for example, it's much more likely to be true.
I think it is quite expected that it is biased to generating output that represent its training data, but this seems like it is not really a solution to the problem. Furthermore, sometimes I want ChatGPT to make things up which is not identical to training data. How do you get it to recognize that it is operating in the realm of fact or not?
I'm not sure larger models with more parameters gets you to where you want to go.
I think many people overstate the problem, I think it is not that serious, but I think a lot of people also try and just dismiss the issue.
Kind of crazy to see GCP only losing $480M in a quarter. I know that sounds ridiculous, but looks like it's on track to actually start posting a profit within a year? That's nice!
Google Cloud also includes Google Workplaces (Faka gSuite, faka Google Apps). That's a high profit area that might be masking GCP's losses. For all we know GCP's losses have remained the same with the growth from Google Workplaces hiding it.
What's your point? When I managed IT I would have considered those 'cloud apps' as opposed to onsite/self hosted, and paying for them would have come out of my OpEx not CapEx budgets. Are they not Cloud?
The point is it paints them as a stronger competitor to AMZN and MSFT when the product categories are different, regardless of how customers do accounting.
But, I feel like the novelty has worn off.