Ask HN: How did Google lose so much ground to OpenAI? Is this really a case of innovator's dilemma + slow moving megacorp? With all the talent they have? |
Ask HN: How did Google lose so much ground to OpenAI? Is this really a case of innovator's dilemma + slow moving megacorp? With all the talent they have? |
Has been since the late 00’s.
Open AI was founded and is staffed by people who see this.
Ultimately I think it is the illusion of a moat (we are too big to fail, we are the default search engine/browser/os etc.) Coupled with over hiring. Along with a fear of taking risk and willingness to disrupt yourself. Ultimately, they became too fearful and focused on building a wall around the kingdom than expanding the kingdom.
Also is it just me, or has the execution from GOOG just been really disappointing? Look at this talk [0] These are not innovators, these are business leaders. Google (and many big tech like AMZN/AWS) have become more packed with these "business leaders" types and not the core, quirky, unusual and strange innovator types. I think it is very much a regression to the "above average" which has killed the ability to launch / productize new technologies (vs. new businesses..)
Compare: https://www.imd.org/research-knowledge/articles/apples-dwind...
> Under Steve Jobs, Apple had a track record of cannibalizing its own products. In 2005, when the demand for the iPod Mini remained huge, the Nano was launched, effectively destroying the revenue stream of an existing product. And while iPod sales were still going through the roof, Jobs launched the iPhone which combined iPod, cell phone, and Internet access into a single device. Three years after the iPhone’s launch, iPad made its debut and raised the prospect of cutting into Mac desktop computer sales. So resolute was Apple’s determination in trading a highly profitable business for an unknown future that Jobs reportedly said “If you don’t cannibalize yourself, someone else will.”
https://www.businessinsider.com/apple-on-cannibalization-201...
> "The iPhone has to become so great that you don't know why you want an iPad," Schiller explained. "The iPad has to be so great that you don't know why you want a notebook. The notebook has to be so great that you don't know why you want a desktop. Each one's job is to compete with the other ones."
> Apple has shed light on this way of thinking before. During a quarterly earnings call with investors in early 2013, Apple CEO Tim Cook said the company's "base philosophy is to never fear cannibalization."
> "If we do, somebody else will just cannibalize it, and so we never fear it," Cook said. "We know that iPhone has cannibalized some iPod business. It doesn’t worry us, but it’s done that. We know that iPad will cannibalize some Macs. That doesn’t worry us."
----
Meanwhile, Google is very cannibalization reluctant to the point of throwing away products that may impact its main revenue sources.
I can't speak from experience, but what I hear from friends of friends is that Azure is really on the come up, though in my experience AWS has still been the default. GCP seemed to be more tempting when they were the only ones throwing tons of money at you and Firebase had less alternatives.
Google's approach to AI advancements follows a similar pattern. Projects like Imagen and Muse boast impressive capabilities but remain locked away, inaccessible to anyone outside of Google. While the company may be a thought leader in the field, the true value of a product lies in its utility. Possessing unique and groundbreaking technology that remains inaccessible to the broader public is comparable to owning an NFT art piece—intriguing but ultimately of limited practical value.
Google has their own LLM and probably access to more data to train future models any other company in the world. What OpenAI seems to have got right was fine tuning their LLM and putting it behind a UI which the average user could use. Then more recently they've created more hype by adding APIs and plugins for their models.
Not saying that what OpenAI's done isn't impressive, but Google really isn't that far behind in terms of tech. 90% of OpenAI's lead is usability stuff.
Google needs to trim some fat from their leadership and streamline their processes if they want to keep up with the agile and innovative pace of OpenAI. It's no wonder they're falling behind if they're more focused on stakeholder alignment sessions than actually developing their AI products. Maybe they should hire some product managers who actually understand technology and engineers who don't overcomplicate things. Otherwise, they'll continue to play catch-up while OpenAI leads the way in the AI industry. Google only wakes up when competition does stuff. See nest for ring, pixel for iPhone or android for iOS.
I ran the exact same prompts through ChatGPT and Bard, and got eerily similar responses. Bard likes bulleted lists better than ChatGPT, but formatting aside the quality of content felt equivalent.
So it seems plausible that Google is behind on productizing and marketing their AI, but I'm not so sure they are behind on the AI itself.
Take for example Gmail, Docs, and any other of their products. The generative stuff exists for a while. Did they implement it into Gmail? Well, yes, a while ago they introduced Compose. Does it generate any text besides super common phrases like "see you tomorrow"? Not reallly. Their leaders are too careful and scared tbh.
Finally, in most big tech companies it's not easy to make major changes without a huge effort and push from leadership. That's because many teams are involved and people move pretty often, so you have huge part of the codebase that even the people who own it don't know very well. So teams prefer not to touch it unless they have to.
This happens frequently in high tech. This is not the 1st, 10th, or 100th time the old guard got left behind in tech.
It seems to me like the truly popular and useful models right now are in the video/image space, which is also not Google’s business.
Google is a behemoth with multiple products and a lot of people with opinions who you have to get through to launch a product.
Also, OpenAI has not unseated Google's dominance in search nor do I see this happening.
but that is nothing compared to how google crafted products then threw them to the graveyard, which did them bad
add a shit demo of bard to that, i dont know anymore, they got massive $$$ to compete in that area but i dont know: lack of vision? lack of people in charge? only insiders in bard could tell
I think neither they know OpenAI expected that this would create such a hype. I mean, GPT 3 has been here for quite a while, and most people just didn’t care.
Elon Musk has been warning about AI for like a decade now. So far, nobody listened. It took a viral chat interface to make people realize what is going on.
Kinda like the internet in 1994 vs. 1999. Or remote work in 2008 vs. 2020. A thing that has been there all along suddenly gets the mainstream’s attention and becomes “cool”.