4 ms·

Google leardship was primarily reacting to threats, such as Amazon S3 and the iPhone, rather than proactively innovating. Despite the fact that AI is widely reg

by tlogan 4y ago

Google leardship was primarily reacting to threats, such as Amazon S3 and the iPhone, rather than proactively innovating. Despite the fact that AI is widely regarded as the future, Google has not made a significant effort to dominate in this field. I think Google needs a visionary leader who can anticipate the future and make strategic investments. I do not think internal structure will change without visionary leader.

skissane 4y ago
> Despite the fact that AI is widely regarded as the future, Google has not made a significant effort to dominate in this field. I think of “Google and AI”, and I remember hearing more about their work on “AI ethics” than work on actual “AI” itself. Of course, ethics is important, but if you are more focused on the ethics of doing something than actually doing it, that’s scarcely a recipe for leading the market-on the contrary, it sounds rather risk-averse-and, maybe that’s unfair, I don’t know, but that’s just the impression I get from my own limited perspective.
- hallqv 4y ago
  [flagged]
  - hef19898 4y ago
    What has "wokeness" to do with any of that?
- FormerBandmate 4y ago
  AI ethics is also big at OpenAI tbf. OpenAI actually releases major models to the public, Google looks like the IBM to their Microsoft right now. Google has created things like TensorFlow tho, OpenAI is flashy but only time will tell if that leads to real, monetizable results
  - amf12 4y ago
    > Google looks like the IBM to their Microsoft right now. I really dislike how Microsoft is being considered to "be better at AI" recently. No, they just integrated another AI solution in Bing. They haven't released any big AI product themselves for a long time.
    - flakiness 4y ago
      The parent comment refers to MS at 2000s, not to the one today.
- guelo 4y ago
  That's just whatever antiwoke media you consume that made a big deal out of the one ethics thing.
  - skissane 4y ago
    From memory, the “antiwoke media” I heard about the “one ethics thing” from is called “Hacker News”
    - guelo 4y ago
      if the shoe fits
- michaelt 4y ago
  > I think of “Google and AI”, and I remember hearing more about their work on “AI ethics” than work on actual “AI” itself. TensorFlow? MobileNet? TPUs? Products like Google Translate, Image Search, Youtube's voice recognition subtitles? Google's done plenty of cutting edge ML stuff. I'll admit their voice assistant is nothing to write home about - but it was the equal of its mediocre competitors at the time it was released.
  - tlogan 4y ago
    Yes: but these seems like engineering products which were not leaders in the market. No clear strategy or vision.
  - fidgewidge 4y ago
    Google uses AI/ML all over the place to incrementally improve their products and always has but what people mean by "AI" in this context is the big LLMs or generative AI.
ethbr0 4y ago
> has not made a significant effort to dominate in this field It's almost like Google is not immune to the fate of being a massive company. And just like prior large, market dominating firms (IBM / Microsoft / Intel), managed to miss a new trend (commodity PC hardware / web/mobile / contract fabbing) due to wildly profitable legacy revenue streams whose sheer size masked competitor acceleration.
amf12 4y ago
I don't think it's the problem with lack of innovation. Google has innovated in a lot of areas - Transformers (the foundation for ChatGPT), Stadia, Project Starline, and some the public may not have heard of yet. The problem lies with translating that innovation to long term product vision, and reimagining how these innovations could be made into a product. I suspect that they announce it to fast too, too soon - before it's proven. Stadia and LaMDA is an example of this. They actually showcased LaMDA last year or the year before that in I/O. The other is that they give up on innovations too soon because they can't think of ways to monetize them.
jocaal 4y ago
> Google has not made a significant effort to dominate in this field. What? Tensorflow, one of the 2 main frameworks for doing AI work is a google project. Google has domain specific processors to accelerate AI workloads - tensor processor. Google Brain. Id argue they have been making the biggest effort out of all the large tech conglomerates. They just haven't made an effort to turn their developments into consumer facing products, instead of milking the ad-cow even more.
- Patrol8394 4y ago
  Google has shown they can build amazing AI that can play chess and go, nice … then what? They built borg and gave k8s to the world and then what? OpenAI took the stage on ChatGPT and AWS dominates the cloud. This goes back to way promotions are handled: clearly people are incentivized to build stuff but no one cares if that would have any path to profitability. Does the world really need fucsia? ChromeOS? A business needs both and needs to incentivize innovation along with long term commitment.
  - jocaal 4y ago
    > and gave k8s to the world and then what? kubernetes came after the rise of aws, likely to keep google in the game > Does the world really need fucsia? ChromeOS? ChromeOS is an extremely popular OS, especially in education markets and there is definitely a need for such an OS. The majority of computers in the world are overkill for what the average end-user is trying to achieve and chromeOS fills that space very neetly. Otherwise I agree with you that technology needs more than just technical vision to succeed, but my opinion is that AI isn't ready for the mainstage and maybe some people within google thought the same and chose to not rush anything. Then google made a massive blunder by going on the defense with their bard unveiling.

by tlogan 4y ago

skissane 4y ago

> Despite the fact that AI is widely regarded as the future, Google has not made a significant effort to dominate in this field. I think of “Google and AI”, and I remember hearing more about their work on “AI ethics” than work on actual “AI” itself. Of course, ethics is important, but if you are more focused on the ethics of doing something than actually doing it, that’s scarcely a recipe for leading the market-on the contrary, it sounds rather risk-averse-and, maybe that’s unfair, I don’t know, but that’s just the impression I get from my own limited perspective.

hallqv 4y ago
[flagged]
- hef19898 4y ago
  What has "wokeness" to do with any of that?
FormerBandmate 4y ago
AI ethics is also big at OpenAI tbf. OpenAI actually releases major models to the public, Google looks like the IBM to their Microsoft right now. Google has created things like TensorFlow tho, OpenAI is flashy but only time will tell if that leads to real, monetizable results
- amf12 4y ago
  > Google looks like the IBM to their Microsoft right now. I really dislike how Microsoft is being considered to "be better at AI" recently. No, they just integrated another AI solution in Bing. They haven't released any big AI product themselves for a long time.
  - flakiness 4y ago
    The parent comment refers to MS at 2000s, not to the one today.
guelo 4y ago
That's just whatever antiwoke media you consume that made a big deal out of the one ethics thing.
- skissane 4y ago
  From memory, the “antiwoke media” I heard about the “one ethics thing” from is called “Hacker News”
  - guelo 4y ago
    if the shoe fits
michaelt 4y ago
> I think of “Google and AI”, and I remember hearing more about their work on “AI ethics” than work on actual “AI” itself. TensorFlow? MobileNet? TPUs? Products like Google Translate, Image Search, Youtube's voice recognition subtitles? Google's done plenty of cutting edge ML stuff. I'll admit their voice assistant is nothing to write home about - but it was the equal of its mediocre competitors at the time it was released.
- tlogan 4y ago
  Yes: but these seems like engineering products which were not leaders in the market. No clear strategy or vision.
- fidgewidge 4y ago
  Google uses AI/ML all over the place to incrementally improve their products and always has but what people mean by "AI" in this context is the big LLMs or generative AI.

ethbr0 4y ago

> has not made a significant effort to dominate in this field It's almost like Google is not immune to the fate of being a massive company. And just like prior large, market dominating firms (IBM / Microsoft / Intel), managed to miss a new trend (commodity PC hardware / web/mobile / contract fabbing) due to wildly profitable legacy revenue streams whose sheer size masked competitor acceleration.

amf12 4y ago

I don't think it's the problem with lack of innovation. Google has innovated in a lot of areas - Transformers (the foundation for ChatGPT), Stadia, Project Starline, and some the public may not have heard of yet. The problem lies with translating that innovation to long term product vision, and reimagining how these innovations could be made into a product. I suspect that they announce it to fast too, too soon - before it's proven. Stadia and LaMDA is an example of this. They actually showcased LaMDA last year or the year before that in I/O. The other is that they give up on innovations too soon because they can't think of ways to monetize them.

jocaal 4y ago

> Google has not made a significant effort to dominate in this field. What? Tensorflow, one of the 2 main frameworks for doing AI work is a google project. Google has domain specific processors to accelerate AI workloads - tensor processor. Google Brain. Id argue they have been making the biggest effort out of all the large tech conglomerates. They just haven't made an effort to turn their developments into consumer facing products, instead of milking the ad-cow even more.

Patrol8394 4y ago
Google has shown they can build amazing AI that can play chess and go, nice … then what? They built borg and gave k8s to the world and then what? OpenAI took the stage on ChatGPT and AWS dominates the cloud. This goes back to way promotions are handled: clearly people are incentivized to build stuff but no one cares if that would have any path to profitability. Does the world really need fucsia? ChromeOS? A business needs both and needs to incentivize innovation along with long term commitment.
- jocaal 4y ago
  > and gave k8s to the world and then what? kubernetes came after the rise of aws, likely to keep google in the game > Does the world really need fucsia? ChromeOS? ChromeOS is an extremely popular OS, especially in education markets and there is definitely a need for such an OS. The majority of computers in the world are overkill for what the average end-user is trying to achieve and chromeOS fills that space very neetly. Otherwise I agree with you that technology needs more than just technical vision to succeed, but my opinion is that AI isn't ready for the mainstage and maybe some people within google thought the same and chose to not rush anything. Then google made a massive blunder by going on the defense with their bard unveiling.