HardCodedBias
2. Youtube meh, like always (weird that it doesn't kill it)
3. Search -- this was the issue. QonQ 0.1% <-- the whole story
Everyone has expected Search to, sooner or later, crack. This looks like it may be it.
If Google had 30B in revenue from coding tools and a few billion from Gemini Enterprise then there would be a story that Google could get through this.
But GDM didn't execute. And that's that.
Who knows the bottom. More than 10-15% seems unlikely. But if search cracks again next quarter then we'll see a real revaluing of the business.
Google did not have the proper focus on AI. They should have owned the space. But there were clearly management issues.
Even if Claude meets all of our wildest goals and earns trillions yearly there is absolutely no “guarantee” that any will flow to shareholders.
1. Shareholders do not have control of the company due to its structure. The Trust controls through class T shares. As a backup the founders control through class F shares.
2. Since the company is a PBC its board is allowed balance dividends with other social benefits as they see fit.
This means that even if Anthropic is wildly successful as shareholders have no general legal claim to compel the company to distribute those profits.
It is absurd that with a 2T IPO upcoming this is not front and center.
Let's apply that throughout the economy. All companies are now bankrupt. All stocks are liquidated and the remaining value (such that it is) is now distributed to lawyers and some to the people.
We've lost our minds.
But given the cost of implementation is low, why not do this!
Good job OAI!
An over time graph with a rolling window of 7d and 28d would likely be a nice addition.
Great work!
We don't know what we don't know. It would be crazy not to ban it, unless it gave truly better performance. It doesn't.
I'm not saying that the US labs (Anthropic, OpenAI, and Meta) aren't pulling their weight, they are all doing amazing work as well, but Deepseek may belong in that tier.
Probably to protect their advertising business which is very strong.
We’ll see how this plays out.
Brilliant.
They are much more than McDonalds.
But indeed your comparison is an apt way to reply to this "article".
It’s 200 layer model.
Great. Good on them for being able to train it.
That gives me hope that there is still areas to improve.
What a bad launch video. Hilarious.
What a powerful model.
The Google brand remains powerful on HN!
I’m shocked.
Our laws in this area are insane : COPPA is very broad. Meta attempted to comply but the state is arguing “willful-blindness” (which makes sense for meta, PII restrictions are very strict). But the lawyers will argue post-hoc that Meta “should have done more” these standards are so vague and argued after the fact that they make a mockery of our system.
Then the state laws being leveraged are crazy broad. California competition law as well as advertising law is being leveraged.
My god, it has to stop.
We’re far past the point of “show me the man, I’ll show you the crime.”
How does he do this? It is simply amazing.
I wonder if GDM has finished their pre-work for their summit on research into how to make Gemini-4 on par with Opus 4.8?
All of the Gemini models have been considerably behind the capabilities frontier. The only exception was 3.0 which seemed quite good, but had latent issues and we were all measuring with the incorrect metric, agentic where it's latent issues were very pronounced.
GDM+Google may have created an exceptionally efficient LLM for serving search. This is likely a great accomplishment (or maybe Google is burning money at a rate unheard of before). But Frontier capability: they have never been in the race.
This is sad, since they had everything necessary to be on or beyond the frontier.
These kind of responses are hilarious.
Someone gives something for free (and indeed this is entirely free) and the top comment is pure complaint.
I don't know why MSL released this, but it is very nice that they did.
GDM management really thought that they were some kind of charity. UNREAL.
I think that is an oxymoron.
I would consider this: maybe useful, do it if you find it entertaining.
But this is mostly marketing, pleasing the sneering class/the elites who believe that simply providing value for others (through sales) is repugnant and beneath them.
It seems that these tools can do real work, and people are paying for that. IMO, that is more than sufficient.
GDM ... not so much.
It's an unbelievable failure.