Is India Ready To Seize The Generative AI Tech Opportunity?
A multi-year multi-billion dollar investment by 微软 into OpenAI not only spruced up the generative AI space but unleashed the reality of an entirely new sector for the planet. Netting over 100 million users in three months not only setup ChatGPT a pole position but clearly set up the US as a leader in emerging technologies, again.
In the wild world of tech, you would be wrong to think there are no intruders -as 谷歌 recently found out. An old foe can stay below the radar for two decades and shake you up with a strategic weapon when they are ready.
Among those that seem rattled are not just 谷歌 but also several Chinese companies 百度 , JD.com, and 阿里巴巴集团 , and others who according to some sources are scrambling to get their versions out. So the covers are off a full-fledged digital arms race for a global competitive edge that now has national importance not just an industry-centric one
When Search had its comparable moment two decades ago perhaps there was not such a nationalistic fervor to it because the US-China engagement was then alive and well. It was heydays of deep collaboration and friendship. All of that changed post-COVID and now anything China is scrutinized and kept at arm's length or outright banned globally.
The deep seeding of Chinese tech into the US over the last two decades is appreciable. China fully exploited the good vibes and made hay when the sun shined. Joint research labs between China and US in biotech, virology, AI science, deep learning, aerospace and avionics, material science, and genetics have both offshore labs as well as embedded within US schools of higher learning.
China's mystery-shrouded policies, closed-door approach to failure analysis, and restricted information flow have alienated it further throwing open doors for alternate economies to jump in filling the supply-chain gaps. India is one of those that can benefit hugely.
Quick Check on India
I looked for any noticeable Indian presence in the generative AI space. Some observations are captured below.
India has missed several industrial revolutions in the past (such as manufacturing, and textile production - downgraded from end product to raw materials in cotton), metals, and toys for example. Over time, India helplessly watched itself becoming a raw material supplier to manufacturing zones like China, Indonesia, Vietnam, and even Bangladesh.
India has a history of jumping in late and picking up the threads and then slowly building up over time by scaling bodies (something we have done appreciably in IT services to about $200 billion annually now after three decades). We have learned significant lessons and changed the game in SaaS products (where our capital efficiency is the best in the world).
We have conceptualized and built the #UPI stack from the ground up and taken it to scale. #UPI now is now the world's largest digital public good in financial payment services, with the highest number of users and transaction volume. India is now in active licensing discussions with multiple governments for deployment into financial systems in those countries.
The Generative AI Landscape
Generative AI refers to machine learning algorithms and models that enable computers to use existing content like text, audio, and video files, images, and even code to create completely new content.
From some of the preliminary research, I did in this space the stack layers are already well formed and have raised fair amounts of funds in the last 3-4 years. Per some estimates, there are over 1200 companies working in this emerging space with over $8-10 billion in funding already been raised. Sources used in this research are mentioned below.
I checked the Gartner 2023 Emerging Tech and Trends Impact Radar where generative AI is mapped under High Mass Impact (indicates how substantial technology will be on existing products and markets). Range ranking (time taken to cross over from early adopter to early majority adopter) is at 3-6 years. But the rate at which it is exploding, it could beat that time frame.
Here is a segmental layer analysis of the generative AI space
Sources: websites a16z, blog sites of Peter Yang, base10, LinkedIn posts Steve Nouri, and others. Some of this data probably needs updating and if you have more authentic data please feel free to add it in the comments section.
Turns text into audio avatars or translates existing avatars' speech
a. Dubverse
b. Basch.io
c. Gaia
Number of Indian companies -3
2. Audio /Video - Audio Editing or Music Generation
Indian companies - None
3. Audio/Video Summarization and Insights
Turns long-form video/audio either into video highlights or text summary of key points
a. Videoverse
b. Magnifi
c. Vidyo
Indian companies -3
4. Audio/Video text to speech and speech-to-text (TTS /STT)
No Indian company
No Indian company
5. Data -Creative organization
AI-assisted brainstorming and idea structuring
No Indian company
6. Data - synthetic data generation
Developing model synthetic data for additional training or research
No Indian company
7. Data-vector search and curation
Comprehension-based identification of related content and information
领英推荐
No Indian company
8. Image -ad collateral
AI-developed images to be used in sales and marketing for ad collateral combines multi-modal
Blend
9. Image Design
AI-generated drawings or plans for graphic design, 3-D architectures
Number of Indian companies - 0
10. Image -Image Editing -
AI-assisted image editing and post-production
Number of Indian companies - 0
10.Image - Text to image - AI-generated stock photos
Number of Indian companies - 0
11. MLOps - workflow tools for model optimization, training
Number of Indian companies - 0
12. Platform - foundational models that enable generative AI applications
Number of Indian companies - 0
13. Text -code generation and documentation
Generating code for app building, querying
Number of Indian companies - 0
14. Text -customer support
Response comprehension for CS and support
Haptik - acquired
Verloop -
Number of Indian companies - 2
15. Text -general writing and editing
use case agnostic text generation and editing for long-form
Longshot
Trials.ai
Number of Indian companies -2
16. Text - marketing and sales (text generation for marketing and sales- blogs, SEO, email marketing)
Peppertype.ai
Scalenut
Number of Indian companies - 2
16. Text -summarization -turning large text or qualitative data into digestible content
TLDR this -
Number of Indian companies - 1
Source: a16z, base10, generative AI blog posts on LinkedIn, blog posts of Peter Yang, Steve Nouri, and others.
There are only a dozen-odd companies at work in this new area at present. It is certainly possible that more companies are already at work but do not fall into these service layers or are not yet in the public eye (disclaimer: this is not a comprehensive list).
We have some experience now creating products (Zoho) and SaaS (several examples), creating product ecosystems, building SI and distribution partners (for specific verticals - enterprises and niche business cases), and scaling both in volume as well as revenues. The funding and capital markets are well established now and within reach. Product discoverability, credibility, and trust factors are also proven beyond doubt. Beyond just creating the necessary technology, we should be able to create leadership positions in these niches, carve out sub-niches and dominate those spaces.
We need to put together an AI industry watch team and take stock of our capabilities here as well. This needs to be done with a sense of urgency if not already. There is no point in being ready when the tide has left the shore.
Clearly, there is much more for us to do. Else we will end up playing catch again.
Free: I am offering a free 30-minute clarity call if you are building a product/service and you need help with strategy, demand creation, and lead generation. To schedule a call: https://calendly.com/tonsepai or DM me on LinkedIn or @tonse_pai on Twitter.
Professor of Practice- Sales and Marketing
1 年Big development. I found your article interesting and useful. Thanks
Dean, T A Pai Management Institute (TAPMI), Bengaluru, MAHE Manipal
1 年Fantastic and detailed Sridhar Tonse Pai . Hopefully we will be able to capture the Generative AI space.