- The Alignment
- Posts
- Bard will feature in Google Search
Happy Monday ! Sundar Pichai has been on a media tour reassuring folks that Google has some big plans for its AI products. Meta on the other hand is releasing new AI capabilities every week ever since the company pivoted its product strategy.
Everyone is part of the AI race, some are playing the long game others are eager to get out in the arena.
As always if you enjoy reading our posts be sure to spread the word !
Google has plans to do all the things that Microsoft has been releasing for the past 3 months (and more). Today however, Google Bard is only open to beta users whereas most of Microsoft's product updates are available to the general public.
But that does not mean that Google’s AI products will forever stay in closed beta testing. In fact Google CEO Sundar Pichai has said that Bard's Chat AI will definitely be making its way into search. There is no public timeline for this, but there is urgency in his voice.
There are two crucial reasons why we haven’t seen Bard out in the open or even tightly integrated with Google search.
Microsoft can make mistakes, Bing is a very small fraction of company revenue. Google can’t make mistakes, the company’s bread and butter is internet search. This is the perfect innovator’s dilemma situation.
If AI and “AGI” is a multi decade pursuit then playing the long game makes sense. Remember, Google was not the first search engine on the market. It was the 27th. First movers don’t always win.
Microsoft’s pace vs Google's patience, two competing strategies, only one winner ?
Meta has released their Segment Anything Model (SAM). An AI model that can identify objects in images and videos. The model works when users either click on elements or prompt the model to select objects and modify them. The model is able to identify objects irrespective of whether it has seen them before.
Image segmentation is a computer vision technique that is used to divide a given image into multiple segments. The main use for such a model comes when the image has to be analysed or processed further. The use cases are wide and varied ranging from dividing web pages into different components, editing images, doing research studies where tacking objects is important and of course in AR/VR.
Meta has also open sourced the two massive datasets it has used to train the model. It’s worth noting that one of them was 400x larger than the previous largest one !
It’s nice to see Meta open sourcing portions of their models for the public to understand. Meta’s AI arm has been shipping and sharing a lot the past couple of months. They’re led by an AI legend and vocal leader in Yann Lecun and you best be sure they will be right in the mix in this AI race.
Around the industry -
Generative AI could transform the way we interact with enterprise software
The ‘AI-obsessed’ photographer who tricked Instagram
‘Mind-reading’ AI: Japan study sparks ethical debate
Generative AI could raise global GDP by 7%
Product Corner -
There are way too many cool products to showcase just one !
Promptly - Build AI apps in minutes
Jarvis - Text your favourite AI assistant with any questions
Text to Video plugin on ChatGPT
AI Intern - With AI Intern skip the grind and focus on the big picture.
Graphmaker - Use chat to create graphs from your data
Is it all a circle of life ?
