13 ms·

Genuine naive question: when it comes to Google HN has generally a negative view of it (pick any random story on Chrome, ads, search, web, working at faang, etc

Genuine naive question: when it comes to Google HN has generally a negative view of it (pick any random story on Chrome, ads, search, web, working at faang, etc. and this should be obvious from the comments), yet when it comes to AI there is a somewhat notable “cheering effect” for Google to win the AI race that goes beyond a conventional appreciation of a healthy competitive landscape, which may appear as a bit of a double standard. Why is this? Is it because OpenAI is seen as such a negative player in this ecosystem that Google “gets a pass on this one”? And bonus question: what do people think will happen to OpenAI if Google wins the race? Do you think they’ll literally just go bust?

antirez 1y ago
Maybe because Google is largely responsible, paying for the research, of most of the results we are seeing now. I'm not a Google fan, in the web side, and in their idea of what software engineering is, but they deserve to win the AI race, because right now all the other players provided a lot less than what Google did as public research. Also, with Gemini 2.5 PRO, there was a big hype moment, because the model is of unseen ability.
- wkat4242 1y ago
  Maybe they deserve it but it would be really bad for the world. Because they will enshittify the hell out of it once they're established. That's their MO. I don't want Google to have a stranglehold over yet another type of online service. So I avoid them. And things are going so fast now, whatever Google has today that might be better than the rest, in two months the rest will have it too. Of course Google will have something new again. But being 2 months behind isn't a huge deal. I don't have to have the 'winning' product. In fact most of my AI tasks go to an 8b llama 3.1 model. It's about on par with gpt 3.5 but that's fine.
  - visarga 1y ago
    The situation with LLMs is much different than search, Google doesn't have such a large lead here. LLMs are social things, they learn from each other, any provider with SOTA model will see its abilities leaked through synthetic training data. That's what GPT-4 did for a year, against the wishes of OpenAI, powering up millions of open model finetunes.
01100011 1y ago
Didn't Google invent the transformer? I think a lot of us see Google as both an evil advertiser and as an innovator. Google winning AI is sort of nostalgic for those of us who once cheered the "Do No Evil"(now mostly "Do Know Evil") company. I also like how Google is making quiet progress while other companies take their latest incremental improvement and promote it as hard as they can.
pkaye 1y ago
I think for a while some people felt the Google AI models are worse but now its getting much better. On the other hand Google has their own hardware so they can drive down the costs of using the models so it keeps pressure on Open AI do remain cost competitive. Then you have Anthropic which has very good models but is very expensive. But I've heard they are working with Amazon to build a data center with Amazons custom AI chips so maybe they can bring down their costs. In the end all these companies will need a good model and lower cost hardware to succeed.
brap 1y ago
I am cheering for the old Google to make a comeback and it seems like the AI race has genuinely sparked something positive inside Google.
wyre 1y ago
Gemini is just that good. From my usage it is much smarter than DeepSeek or Claude 3.7 Thinking models. A lot of Google’s market share across its services comes from the monopoly effects Google has. The quality of Gemini 2.5 is noticeably smarter than its competitors so I see the applause for the quality of the LLM and not for Google. I think it’s way too early to say anything about who is winning the race. There is still a long way to go; o3 scores highest in Humanity’s Last Exam (https://agi.safe.ai/ https://agi.safe.ai/) at 20%, 2.5 scores 18%.
sothatsit 1y ago
2.5 Pro is free, and I'm sure there's a lot of people who have just never tried the best models because they don't want to pay for them. So 2.5 Pro probably blows their socks off. Whereas, if you've been paying for access to the best models from OpenAI and Anthropic all along, 2.5 Pro doesn't feel like such a drastic step-change. But going from free models to 2.5 Pro is a crazy difference. I also think this is why DeepSeek got so much attention so quickly - because it was free.
julianeon 1y ago
It's been a while since they won something the "old" Google way: by building a superior product that is #1 on its merits. In that sense Gemini is a throwback: there's no trick - it's objectively better than everything else.
sagarpatil 1y ago
Most of us weren’t using Gemini pro models (1.0, 1.5, 2.0) but the recent 2.5 pro is such a huge step up. It’s better than 3.7 sonnet for coding. Better than o1, o3-mini models and now o3 and o4-mini. It’s become my daily driver. It does everything I need with almost 100% accuracy, is cheap, fast, 1 million context window, uses google web search for grounding, can fetch YouTube video transcripts, can fetch website content, works in google workspace: Gmail, Docs, Sheets. Really hard to beat this combo. Oh and if you subscribe to their AI plan it comes with 2 TB drive storage.
oezi 1y ago
The key is Gemini being free through AI Studio. This makes their technical improvement more impressive when OpenAI sells their best models at ridiculous prices. If Google engages in price dumping as a monopolist remains to be seen but it feels like it. The LLM race is fast paced and no moat has developed. People are switching on a whim if better models (by some margin) show up. When will OpenAI, Anthropic or DeepSeek counter 2.5 Pro? And will it be before Google releases the next Pro? OpenAI commands a large chunk of the consumer market and they have considerable funds after their last round. They won't fold this or next year. If Google wants to win this they must come up with a product strategy integrating their search business without seriously damaging their existing search business to much. This is hard.
int_19h 1y ago
I dislike Google rather strongly due to their ad-based business model, and I was previously very skeptical of their AI offerings because of very lackluster performance compared to OpenAI and Claude. But I can't help but be impressed with Gemini Pro 2.5 for "deep research" and agentic coding. I have subscriptions with all three so that I can keep up with SOTA, but if I had to choose only one to keep, right now it'd be Gemini. That said I still don't "cheer" for them and I would really rather someone else win the race. But that is orthogonal to recognition of observed objective superiority.
greentea23 1y ago
I prefer OpenAI and Anthropic big time because they are fresh players with less dominance over other aspects of digital life. Not having to login to an insidious tracker like Google is worth significantly worse performance. Although I have little FOMO here avoiding Gemini because evaluating these models on real world use cases remains quite subjective imo.
jonas21 1y ago
A lot of the negativity toward Google stems from the fact that they're the big, dominant player in search, ads, browsers, etc., rather than anything that they've done or any particular attribute of the company. In AI, they're still seen as being behind OpenAI and others, so we don't see the same level of negativity.
summerlight 1y ago
Because now it has brought real competitions to the field. GPT was the king and Claude had been the only meaningful challenger for a while but OpenAI didn't care about Anthropic but just be obsessed with Google. Gemini took a quite time to set the pipeline so initial version was not enough to push the frontier; you remember the days when Google released a new model, OpenAI just responded with some old models in their silo within a day only to crush them. That does not happen anymore and they're forced to develop a better model.
CephalopodMD 1y ago
As a googler working in LLM space, this feels like revisionist history to me haha! I remember a completely different environment only a few months ago when Anthropic was the darling child, and before that it was OpenAI (and for like 4 weeks somewhere in there, it was Deepseek). For literally years at this point, every time Bard or Gemini would make a major release, it would be largely ignored or put down in favor of the next "big thing" OpenAI was doing or Claude saturating coding benchmarks, never mind that Google was often just behind with the exact same tech ready to go, in some cases only missing their demo release by literally 1 day (remember live voice?). And every time this happened, folks would be posting things to the effect of "LOL I can't believe Google is losing the AI race - didn't they invent this?", "this is like Microsoft dropping the ball on mobile", "Google is getting their lunch eaten by scrappy upstarts," etc. I can't lie, it stings a bit when that's what you work on all day. 2.5 was quite good. Not stupidly good like the jump from GPT 2 to 3 or 3.5 to 4, but really good. It was a big jump in ELO and benchmarks. People like it, and I think it's just psychologically satisfying that the player everybody would have expected to win the AI race is currently in the lead. Gemini finally gets a day in the sun. I'm sure this will change with whenever somebody comes up with the next big idea though. It probably won't take much to beat Gemini in the long run. There is literally zero moat.

antirez 1y ago

Maybe because Google is largely responsible, paying for the research, of most of the results we are seeing now. I'm not a Google fan, in the web side, and in their idea of what software engineering is, but they deserve to win the AI race, because right now all the other players provided a lot less than what Google did as public research. Also, with Gemini 2.5 PRO, there was a big hype moment, because the model is of unseen ability.

wkat4242 1y ago
Maybe they deserve it but it would be really bad for the world. Because they will enshittify the hell out of it once they're established. That's their MO. I don't want Google to have a stranglehold over yet another type of online service. So I avoid them. And things are going so fast now, whatever Google has today that might be better than the rest, in two months the rest will have it too. Of course Google will have something new again. But being 2 months behind isn't a huge deal. I don't have to have the 'winning' product. In fact most of my AI tasks go to an 8b llama 3.1 model. It's about on par with gpt 3.5 but that's fine.
- visarga 1y ago
  The situation with LLMs is much different than search, Google doesn't have such a large lead here. LLMs are social things, they learn from each other, any provider with SOTA model will see its abilities leaked through synthetic training data. That's what GPT-4 did for a year, against the wishes of OpenAI, powering up millions of open model finetunes.

01100011 1y ago

Didn't Google invent the transformer? I think a lot of us see Google as both an evil advertiser and as an innovator. Google winning AI is sort of nostalgic for those of us who once cheered the "Do No Evil"(now mostly "Do Know Evil") company. I also like how Google is making quiet progress while other companies take their latest incremental improvement and promote it as hard as they can.

pkaye 1y ago

I think for a while some people felt the Google AI models are worse but now its getting much better. On the other hand Google has their own hardware so they can drive down the costs of using the models so it keeps pressure on Open AI do remain cost competitive. Then you have Anthropic which has very good models but is very expensive. But I've heard they are working with Amazon to build a data center with Amazons custom AI chips so maybe they can bring down their costs. In the end all these companies will need a good model and lower cost hardware to succeed.

brap 1y ago

I am cheering for the old Google to make a comeback and it seems like the AI race has genuinely sparked something positive inside Google.

wyre 1y ago

Gemini is just that good. From my usage it is much smarter than DeepSeek or Claude 3.7 Thinking models. A lot of Google’s market share across its services comes from the monopoly effects Google has. The quality of Gemini 2.5 is noticeably smarter than its competitors so I see the applause for the quality of the LLM and not for Google. I think it’s way too early to say anything about who is winning the race. There is still a long way to go; o3 scores highest in Humanity’s Last Exam (https://agi.safe.ai/ https://agi.safe.ai/) at 20%, 2.5 scores 18%.

sothatsit 1y ago

2.5 Pro is free, and I'm sure there's a lot of people who have just never tried the best models because they don't want to pay for them. So 2.5 Pro probably blows their socks off. Whereas, if you've been paying for access to the best models from OpenAI and Anthropic all along, 2.5 Pro doesn't feel like such a drastic step-change. But going from free models to 2.5 Pro is a crazy difference. I also think this is why DeepSeek got so much attention so quickly - because it was free.

julianeon 1y ago

It's been a while since they won something the "old" Google way: by building a superior product that is #1 on its merits. In that sense Gemini is a throwback: there's no trick - it's objectively better than everything else.

sagarpatil 1y ago

Most of us weren’t using Gemini pro models (1.0, 1.5, 2.0) but the recent 2.5 pro is such a huge step up. It’s better than 3.7 sonnet for coding. Better than o1, o3-mini models and now o3 and o4-mini. It’s become my daily driver. It does everything I need with almost 100% accuracy, is cheap, fast, 1 million context window, uses google web search for grounding, can fetch YouTube video transcripts, can fetch website content, works in google workspace: Gmail, Docs, Sheets. Really hard to beat this combo. Oh and if you subscribe to their AI plan it comes with 2 TB drive storage.

oezi 1y ago

The key is Gemini being free through AI Studio. This makes their technical improvement more impressive when OpenAI sells their best models at ridiculous prices. If Google engages in price dumping as a monopolist remains to be seen but it feels like it. The LLM race is fast paced and no moat has developed. People are switching on a whim if better models (by some margin) show up. When will OpenAI, Anthropic or DeepSeek counter 2.5 Pro? And will it be before Google releases the next Pro? OpenAI commands a large chunk of the consumer market and they have considerable funds after their last round. They won't fold this or next year. If Google wants to win this they must come up with a product strategy integrating their search business without seriously damaging their existing search business to much. This is hard.

int_19h 1y ago

I dislike Google rather strongly due to their ad-based business model, and I was previously very skeptical of their AI offerings because of very lackluster performance compared to OpenAI and Claude. But I can't help but be impressed with Gemini Pro 2.5 for "deep research" and agentic coding. I have subscriptions with all three so that I can keep up with SOTA, but if I had to choose only one to keep, right now it'd be Gemini. That said I still don't "cheer" for them and I would really rather someone else win the race. But that is orthogonal to recognition of observed objective superiority.

greentea23 1y ago

I prefer OpenAI and Anthropic big time because they are fresh players with less dominance over other aspects of digital life. Not having to login to an insidious tracker like Google is worth significantly worse performance. Although I have little FOMO here avoiding Gemini because evaluating these models on real world use cases remains quite subjective imo.

jonas21 1y ago

A lot of the negativity toward Google stems from the fact that they're the big, dominant player in search, ads, browsers, etc., rather than anything that they've done or any particular attribute of the company. In AI, they're still seen as being behind OpenAI and others, so we don't see the same level of negativity.

summerlight 1y ago

Because now it has brought real competitions to the field. GPT was the king and Claude had been the only meaningful challenger for a while but OpenAI didn't care about Anthropic but just be obsessed with Google. Gemini took a quite time to set the pipeline so initial version was not enough to push the frontier; you remember the days when Google released a new model, OpenAI just responded with some old models in their silo within a day only to crush them. That does not happen anymore and they're forced to develop a better model.

CephalopodMD 1y ago

As a googler working in LLM space, this feels like revisionist history to me haha! I remember a completely different environment only a few months ago when Anthropic was the darling child, and before that it was OpenAI (and for like 4 weeks somewhere in there, it was Deepseek). For literally years at this point, every time Bard or Gemini would make a major release, it would be largely ignored or put down in favor of the next "big thing" OpenAI was doing or Claude saturating coding benchmarks, never mind that Google was often just behind with the exact same tech ready to go, in some cases only missing their demo release by literally 1 day (remember live voice?). And every time this happened, folks would be posting things to the effect of "LOL I can't believe Google is losing the AI race - didn't they invent this?", "this is like Microsoft dropping the ball on mobile", "Google is getting their lunch eaten by scrappy upstarts," etc. I can't lie, it stings a bit when that's what you work on all day. 2.5 was quite good. Not stupidly good like the jump from GPT 2 to 3 or 3.5 to 4, but really good. It was a big jump in ELO and benchmarks. People like it, and I think it's just psychologically satisfying that the player everybody would have expected to win the AI race is currently in the lead. Gemini finally gets a day in the sun. I'm sure this will change with whenever somebody comes up with the next big idea though. It probably won't take much to beat Gemini in the long run. There is literally zero moat.