Google to Accelerate 20 Product Launches in Response to ChatGPT
Google to Accelerate 20 Product Launches in Response to ChatGPT
Pichai's "Green Lane" initiative targets processes that slow commercialization
The New York Times reported last week that Google is indeed changing its plans around AI-based products in response to the enormous popularity of OpenAI’s ChatGPT. News that Microsoft just completed a new multibillion-dollar investment in OpenAI and became its exclusive cloud provider is sure to put even more momentum behind these efforts. Nico Grant of the New York Times reported:
“Mr. Pichai declared a “code red,” upending existing plans and jump-starting A.I. development. Google now intends to unveil more than 20 new products and demonstrate a version of its search engine with chatbot features this year, according to a slide presentation reviewed by The New York Times and two people with knowledge of the plans who were not authorized to discuss them…
“Mr. Pichai has tried to accelerate product approval reviews, according to the presentation reviewed by The Times. The company established a fast-track review process called the ‘Green Lane’ initiative, pushing groups of employees who try to ensure that technology is fair and ethical to more quickly approve its upcoming A.I. technology…
“The company will also find ways for teams developing A.I. to conduct their own reviews, and it will ‘recalibrate’ the level of risk it is willing to take when releasing the technology, according to the presentation…
“Google listed copyright, privacy and antitrust as the primary risks of the technology in the slide presentation. It said that actions, such as filtering answers to weed out copyrighted material and stopping A.I. from sharing personally identifiable information, are needed to reduce those risks…
“For the chatbot search demonstration that Google plans for this year, getting facts right, ensuring safety and getting rid of misinformation are priorities. For other upcoming services and products, the company has a lower bar and will try to curb issues relating to hate and toxicity, danger and misinformation rather than preventing them, according to the presentation.”
Code Red and Rightfully So
This “code red” language has now shown up in enough places that it seems likely that Google is indeed planning a response to OpenAI’s rapid ascent. Add to this that the New York Times reporters reviewed an internal presentation and spoke with some insiders, and the likelihood of this all being true seems high.
Why all of the scrambling? Google went from a big mindshare leader in AI to a lagging competitor over the past six months. It is not just ChatGPT, though that was clearly the catalyst to increase Google’s sense of urgency. The planned introduction of conversational search by existing competitors such as Microsoft’s Bing and new rivals like Perplexity AI raises the stakes further. Google’s core economic engine is under threat.
Google Imagen looks like an amazing text-to-image generator. If it comes out in May 2023, it will be a full year behind OpenAI’s DALL-E 2 for market introduction and nine months behind Midjourney and Stable Diffusion. OpenAI introduced GPT-3 APIs for developers in 2020. Will Google be three years behind in that segment?
Will Google Wait Until I/O?
Google generally likes to wait to introduce new products based on technology advances until its annual I/O developer conference. However, that is typically in May. Will the company wait another four months to step forward into the news and product cycles? I think it will, but we may see one or two new applications in a preview mode before that. The “Code Red” indicates this is now possible.
The likelihood of a general availability release of any of the AI-based products in Google’s portfolio this quarter is low. There are too many considerations that Google must take into account, and the bureaucracy only bends so far in the face of exhortations to “speed it up.”
Google LaMDA is already available in limited release through Test Kitchen. This might be a good candidate for a wider release. The downside now is that LaMDA’s first release is likely to be limited to lists and creative writing, and its current version is not going to show well against ChatGPT. That could further erode Google’s mindshare in this important new segment.
Imagen is a better bet for early release as it is very far along in development, and it is harder to make head-to-head comparisons between image generators. Google could still be first with a general availability text-to-video solution, but it seems likely that OpenAI or Meta will beat it to market. All three companies say they are holding back on releasing these technologies while they sort out trust and safety features to limit the risk of inappropriate content generation. Google may face the biggest risk in this regard due to regulatory scrutiny, which may, in turn, further slow its rollout.
The code-completion solution PaLM-Coder 2 might be a dark horse for early release. GitHub Copilot, based on OpenAI’s Codex, looks like a smash hit. Google would surely like to have a competitive offering in the market soon, and trust and safety risks are less likely. Then again, there probably are IP infringement risks that Google will want to avoid. I suspect this will be announced at Google I/O in front of a crowd of supportive Android developers.
The other products, such as an automated green screen image maker, a wallpaper maker for Pixel smartphones, and a visualization for three-dimensional shoes, seem less promising. If Google wants to recapture some AI market mindshare, it will have to go bigger. The tool that summarizes “videos by generating a new one” may be promising, but this is also the type of product that Google often fails to deliver on key user requirements.
The Generative AI Wars Have Begun
All of this is occurring amidst a backdrop of 12,000 layoffs at Google. That is not likely to impact Google’s AI group in any significant way. Sundar Pichai’s email to employees about the pending layoffs mentioned several times how important AI was to the company’s future.
“I am confident about the huge opportunity in front of us thanks to the strength of our mission, the value of our products and services, and our early investments in AI. To fully capture it, we’ll need to make tough choices…
“Being constrained in some areas allows us to bet big on others. Pivoting the company to be AI-first years ago led to groundbreaking advances across our businesses and the whole industry.
”Thanks to those early investments, Google’s products are better than ever. And we’re getting ready to share some entirely new experiences for users, developers and businesses, too. We have a substantial opportunity in front of us with AI across our products and are prepared to approach it boldly and responsibly…”
There you have it. The New York Times and Sundar Pichai both indicate that new AI-based products are coming. This was expected, but the ChatGPT phenomenon is changing everyone’s calculus about what the window of opportunity looks like. Microsoft’s deeper ties with OpenAI will add new uncertainty.
The technology companies behind the most promising generative AI advances seemed almost collegial with their competitors just a few months ago. That was when everyone thought we were still in the research phase of the industry. We are now entering the commercialization phase, which means fierce competition for revenue and market share to sustain further technology and product advances.
The rapid adoption of text-to-image generators such as DALL-E, Midjourney, and Stable Diffusion hinted in 2022 that the market was moving more quickly than many anticipated. The early success of GitHub Copilot reinforced this impression. Then ChatGPT blew everything up and changed assumptions almost overnight. Microsoft piled more onto this shift by announcing the integration of OpenAI technology across its product stack.
Google is in reaction mode at the moment. However, it has proven to be the best fast-follower of technology trends in recent memory. It’s time to see if the company can still live up to that track record or if it will become the next high-profile victim of the innovator’s dilemma.
