Why did Google didn’t create ChatGPT? | Hacker News Reader

Why did Google didn’t create ChatGPT?

3 pointsalfor7 comments

It seem to me that Google had plenty of time and the expertise to create something like ChatGPT. I wonder how could they sleep on that technology for so long, they are now at the fourth generation of TPU, they created TensorFlow that was the leading library of machine learning before PyTorch, they created the transformer architecture, they had all the computing and data.

How could a company with such massive ressource get outdone by a startup, well it seems it always go that way.

View on HN

Probably because they knew about hallucinations and didn’t think the tech was ready for public consumption. TBH it really isn’t quite ready yet, but OpenAI forced them to release something.

Google is at the top of the industry, making billions, there is no point in disrupting the market, without a clear business model

Now the little guy, if they can throw a curve ball, they will probably increase their market share

Happens all the time, employee works in Big co, sees technology not being taken advantage, opens startup, open source part of business for competitive advantage

They own search, they own ads, and any disruption from chat-style agents disrupts those hard. So, as long as no one else had it in a compelling, public-facing product, stalling was in their interests.

(Not quite sure why their hair on fire wasn’t until the Bing/GPT-4 release; ChatGPT should have been enough.)

I asked Bard Why did Google didn't create ChatGPT:

Google did not create ChatGPT because it is a text-generation chatbot developed by OpenAI. Google has its own text-generation chatbot called LaMDA, which is still in research mode.

Google is a company that is known for its innovation and its willingness to take risks. However, the company is also very careful about its reputation and its customer relationships. Google does not want to release a product that is not ready for prime time, and it does not want to damage its reputation by releasing a product that is not accurate or reliable.

ChatGPT is a very impressive piece of technology, but it is still in its early stages of development. There are some concerns about the accuracy of the chatbot, and there are some concerns about the potential for the chatbot to be used to spread misinformation or to generate harmful content. Google is not willing to take these risks, and so it has decided not to release a ChatGPT competitor.

Google is likely to continue to develop its own text-generation chatbot, and it is likely to release a more polished and more reliable product in the future. However, for now, Google is content to let OpenAI take the lead in the development of chatbot technology.

If you haven't been following, they have comparable models, they just haven't made them publicly available. Bard is now in beta, the GCP APIs are on the horizon

See PaLM and CaLM for some recent papers.

I know, I find it strange that they didn’t make something better way sooner considering the ressource they had: scientists, computing, data, money.

It seem to me they prefered to keep it closed, away from the public, the questions, the risks.

If Google has AGI, or just basic AI with all the data they have on everyone, it’s a giant spying machine on much of the world and with a monopoly on web advertizing.

I think they had the golden goose and didn’t want to risk anything, just keep it going as long as possible.

What do you think? /Tinfoil hat off.

I think they are being more cautious than rushing a chatbot that has hallucinations. They do have access to better real time data, Bard is in beta

I think we shouldn't jump to conclusions, things are just getting started. Revisit the question in 3 years

Reply on news.ycombinator.com