Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google is winning on every AI front

thealgorithmicbridge.com

Post reply on HN
Print view View original

531–540 of 851 posts

Re: Google is winning on every AI front

#531

Post by dcow »

jnwatson wrote: post #456
dcow wrote: post #367

Earlier quoted context omitted.

I don’t think “AI” as a market is “winner-takes-anything”. Seriously. AI is not a product, it’s a tool for building other products. The winners will be other businesses that use AI tooling to make better products. Does OpenAI really make sense as a chatbot company?

AI is a product when you slap an API on top and host it for other businesses to figure out a use case. In a gold rush, the folks that sell pickaxes make a reliable living.

Nvidia is selling the shovels.

Re: Google is winning on every AI front

#532

Post by mikeocool »

jnwatson wrote: post #456
dcow wrote: post #367

Earlier quoted context omitted.

I don’t think “AI” as a market is “winner-takes-anything”. Seriously. AI is not a product, it’s a tool for building other products. The winners will be other businesses that use AI tooling to make better products. Does OpenAI really make sense as a chatbot company?

AI is a product when you slap an API on top and host it for other businesses to figure out a use case. In a gold rush, the folks that sell pickaxes make a reliable living.

Basically every tech company likes to say they are selling pickaxes, but basically no VC funded company matches that model. To actually come out ahead selling pickaxes you had to pocket a profit on each one you sold.

If you sell your pickaxes at a loss to gain market share, or pour all of your revenue into rapid pickaxe store expansion, you’re going to be just as broke as prospectors when the boom goes bust.

Re: Google is winning on every AI front

#533

Post by mike_hearn »

foota wrote: post #344
mike_hearn wrote: post #294

Earlier quoted context omitted.

TPUs aren't necessarily a pro. They go back 15 years and don't seem to have yielded any kind of durable advantage. Developing them is expensive but their architecture was often over-fit to yesterday's algorithms which is why they've been through so many redesigns. Their competitors have routinely moved much faster using CUDA. Once the space settles down, the balance might tip towards specialized accelerators but NVID…

Haven't Nvidia published roughly as many chip designs in the same period?

The issue isn't number of designs but architectural stability. NVIDIA's chips have been general purpose for a long time. They get faster and more powerful but CUDA has always been able to run any kind of neural network. TPUs used to be over-specialised to specific NN types and couldn't handle even quite small evolutions in algorithm design whereas NVIDIA cards could. Google has used a lot of GPU hardware too, as a consequence.

Re: Google is winning on every AI front

#534

Post by anon84873628 »

netcan wrote: post #435
panarky wrote: post #430

Earlier quoted context omitted.

> chatbots ... provided for free ... ads Just because the first LLM product people paid for was a chatbot does not mean that chat will be the dominant commercial use of AI. And if the dominant use is agents that replace knowledge workers, then they'll cost closer to $2000 per month than $20 or free, and an ad-based business model won't work.

True. This is my point too. The actual business models and revenue sources are still unknown. Consumer subscriptions happens to be the first major model. Ads still aren't. Many other models could dwarf either of these. It's very early to call the final score.

I still think it's pretty clear. Google doesn't have to get a new business off the ground, just keep improving the integration into Workspace, Gmail, Cloud, Android etc. I don't see users paying for ChatGPT and then copy/pasting into those other places even if the model is slightly better. Google will just slowly roll out premium plans that include access to AI features.

And as far as selling pickaxes go, GCP is in a far better position to serve the top of market than OpenAI. Some companies will wire together multiple point solutions but large enterprises will want a consolidated complete stack. GCP already offers you compute clusters and BigQuery and all the rest.

Re: Google is winning on every AI front

#535

Post by mike_hearn »

alienthrowaway wrote: post #523
mike_hearn wrote: post #294

Earlier quoted context omitted.

TPUs aren't necessarily a pro. They go back 15 years and don't seem to have yielded any kind of durable advantage. Developing them is expensive but their architecture was often over-fit to yesterday's algorithms which is why they've been through so many redesigns. Their competitors have routinely moved much faster using CUDA. Once the space settles down, the balance might tip towards specialized accelerators but NVID…

> Developing them is expensive So are the electric and cooling costs at Google's scale. Improving perf-per-watt efficiency can pay for itself. The fact that they keep iterating on it suggests it's not a negative-return exercise.

TPUs probably can pay for themselves, especially given NVIDIA's huge margins. But it's not a given that it's so just because they fund it. When I worked there Google routinely funded all kinds of things without even the foggiest idea of whether it was profitable or not. There was just a really strong philosophical commitment to doing everything in house no matter what.

Re: Google is winning on every AI front

#536

Post by kortilla »

brookst wrote: post #452
istjohn wrote: post #419

Earlier quoted context omitted.

> No guarantees that Google discover a meaningful business model here... I don't understand this sentiment at all. The business model writes itself (so to speak). This is the company that perfected the art of serving up micro-targeted ads to people at the moment they are seeking a solution to a problem. Just swap the search box for a chat bot. For a while they'll keep the ads off to the side, but over time the ads wi…

I don’t think it works. Search is the perfect place for ads for exactly the reasons you state: people have high intent. But a majority of chatbot usage is not searching for the solution to a problem. And if he Chatbot is serving the ads when I’m using it for creative writing, reformatting text, having a python function, written, etc, I’m going to be annoyed and switch to a different product. Search is all about infor…

The main usage of chatgpt I’ve seen amongst non-programmers is a direct search replacement with tons of opportunity for ads.

People ask for recipes, how to fix things around the house, for trip itinerary ideas, etc.

Re: Google is winning on every AI front

#537

Post by wkat4242 »

wkat4242 wrote: post #453
Aurornis wrote: post #381

Earlier quoted context omitted.

> They pay the drivers a pittance. My ex drove Uber for a while and it wasn't really worth it. I keep hearing this online, but every time I’ve used an Uber recently it’s driven by someone who says they’ve been doing it for a very long time. Seems clear to me that it is worth it for some, but not worth it if you have other better job options or don’t need the marginal income.

Maybe it differs per country. This was in Spain.

PS: I know that in Romania it's the opposite. Uber is kinda like a luxury taxi there. Normal taxis have standard rates, but these days it's hardly enough to cover rising fuel prices. So cars are ancient and un a bad state of repair, drivers often trick foreigners. A colleague was even robbed by one. Uber is much more expensive but much safer (and still cheap by western standards).

Re: Google is winning on every AI front

#538

Post by mike_hearn »

HarHarVeryFunny wrote: post #324
mike_hearn wrote: post #227

Earlier quoted context omitted.

You're assuming ads would be subtly worked into the answers. There's no reason it has to be done that way. You can also have a classic text ads system that's matching on the contents of the discussions, or which triggers only for clearly commercial queries "chatgpt I want to eat out tonight, recommend me somewhere", and which emits visually distinct ads. Most advertisers wouldn't want LLMs to make fake recommendation…

There are lots of ways that advertising could be tied to personal interests gleaned by having access to someone's ChatBot history. You wouldn't necessarily need to integrate advertisements into the ChatBot itself - just use it as a data gathering mechanism to learn more about the user so that you can sell that data and/or use it to serve targetted advertisements elsewhere. I think a big commercial opportunity for Cha…

Right, referral fees is pay-per-click advertising.

The obvious way to integrate advertising is for the LLM to have a tool to search an ad database and display the results. So if you do a commercial query the LLM goes off and searches for some relevant ads using everything it knows about you and the conversation, the ad search engine ranks and returns them, the LLM reads the ad copy and then picks a few before embedding them into the HTML with some special React tags. It can give its own opinion to push along people who are overwhelmed by choice. And then when the user clicks an ad the business pays for that click (referral fee).

Re: Google is winning on every AI front

#539

Post by ranyume »

thfuran wrote: post #516
ranyume wrote: post #396

Earlier quoted context omitted.

> Refusing to answer their questions doesn't squash that, it leads them to potentially less trustworthy sources. But that's good

For who?

For the reader.

The AI won't tell the reader what to think in an authoritative voice. This is better than the AI trying to decide what is true and what isn't.

However, the AI should be able to search the web and present it's findings without refusals. Obviously, always presenting the sources. And the AI should never use an authoritative tone and it should be transparent about the steps it took to gather the information, and present the sites and tracks it didn't follow.

Re: Google is winning on every AI front

#540

Post by anon84873628 »

tsimionescu wrote: post #481
UltraSane wrote: post #473

Earlier quoted context omitted.

LLM based advertising has amazing potential when you consider that you can train them to try to persuade people to buy the advertised products and services.

That seems like a recipe for class action false advertising lawsuits. The AI is extremely likely to make materially false claims, and if this text is an advertisement, whoever placed it is legally liable for that.

I don't think we should expect that risk to dissuade these companies. They will plow ahead, fight for years in court, then slightly change the product if forced to ¯\_(ツ)_/¯
Post reply on HN