Is Generative A.I. a New Paradigm?
Is there really a 13-year cycle in consumer tech and adoption?
Hey Guys,
Whether you are into code, a data engineer or into machine learning, Generative A.I. is catching a lot of attention.
We’ll know more as GPT-4 is announced.
I got an Email from Peter Diamandis and it got me thinking. I’m no Singularity University graduate, but is all the stories out of Silicon Valley really serious on this?
Peter like many among our seniors are saying something to the effect of the following:
It’s (advent of Generative A.I.) is a moment that reminds me of a number of disruptive and highly opportunistic periods:
The birth of the internet
The adoption of smartphones
The launch of cloud services
Is Generative A.I. that big of a deal?
The integration of A.I. into all businesses!
It’s hard to really decide on dates when these events occurred, but let’s just say:
Internet arrived to the masses in 1994.
The adoption of smartphones. First iPhone was in 2007
Smartphone adoption started to really accelerated in 2013. This is when Global smartphone sales surpassed the sales figures for feature phones in early 2013.
AWS was founded in 2006, but it took at least until 2016 until the world realized how important and big Cloud services were going to become.
While Generative A.I. has been around a while with GNNs, Transformer came out in 2017.
While it feels like all of these things have “been around forever”, the pace of technology in our lives is actually relatively fast.
If we consider the Cloud a core technology story and not a consumer trend, then the 13-year cycle is thus:
1994 - Birth of Internet
2007 - Mobile phones
2020 - Birth of Generative A.I. (or Cloud Computing)
2033 - ? (Brain Computer Interface, BCI)
I think the 13-year cycle makes some sense, some might even argue that 2033 would finally mean some the Metaverse is ready for mass consumption, or will it be Quantum computing that are scalable and allow for a legit brain-computer-interface BCI?
If these are true, I’d suggest that 2020 could also be replaced with how important Cloud computing became. Generative A.I. is a flashy feature, but perhaps not a consumer movement of mass adoption.
Silicon Valley and Venture Capitalists are saying:
We are at a VERY unique moment, on the cusp of a revolution that will usher in a number of trillion-dollar companies/industries.
A.I is Becoming Ubiquitous
Silicon Valley lore says A.I. is the prime mover.
Business models are about to change across all industries, and the technology that is driving this is the widespread use of artificial intelligence everywhere.
If this is true learning software development, data science, machine learning and related Cloud skills is like participating in a new world.
In 2022, that almost feels like history, let’s recall that on social media platforms such as Twitter and Reddit are filled with images created by generative machine learning models such as DALL-E and Stable Diffusion. TikTok also had time-travellers posting, but that’s another story!
For posterity however, let’s be impressed with the buzz Generative A.I. has had in 2022, where in just a few months GPT-3 matured, DALL-E 2 came out and so did so many startups.
Stability.AI in August really was peak hype. However Venture Capitalists hopped on the bandwagon of Generative A.I. ever since to try to offset the tech layoffs and the weird climate of uncertainty for 2023.
Generative AI is not new. With a few notable exceptions, most of the technologies we’re seeing today have existed for several years. Still in the back of our minds lingers the A.I. dogma: it basically goes like this:
How should you think about this? Here are a few summaries:
“Artificial intelligence could have more profound implications for humanity than electricity or fire.” - Sundar Pichai, CEO, Alphabet
“Companies have to race to build AI or they will be made uncompetitive. Essentially, if your competitor is racing to build AI, they will crush you." - Elon Musk
I summarize it like this: “There will be two kinds of companies at the end of this decade... Those that are fully utilizing AI, and those that are out of business.”
It’s a bit like a simultaneous pitch for Google, Tesla and the Cloud as a whole (Azure, AWS, Google Cloud), etc…
That’s nice, but that doesn’t mean Generative A.I. is a new paradigm or that companies are even very efficient at being adopting machine learning, nevermind MLops or a functional modern data stack.
Peter Diamandis sounds a bit like an unhinged Cathie Wood claiming that A.I. is an exponential force of nature that we should merge ourselves with the machine immediately! It’s like Jack Dorsey in the middle of a Bitcoin dopamine hit suddenly deciding to call Square (which was a perfectly good name) Block instead. I wonder how that will turn out as the crypto winter is looking bleak?
A.I. Never Rests
But A.I. hype never rests because A.I. research, A.I. collectives and access to compute keeps getting cheaper. A.I. is starting to have a mind and momentum of its own embedded in Capitalism, from China to India to America.
It’s perhaps the young demographics of India and talent in A.I. that may be one of the brighter stories of the 2020s. This would be a good time to found a Generative A.I. startup in India. As China struggles with real-estate crisis, zero-covid and politics and America still pretends it’s number one, India has a real window of opportunity in the 2023 to 2033 period.
Generative A.I. is a Diversification and Digital Transformation of A.I.
Generative A.I. is certainly not the birth of a new internet or a consumer device, but it is something fundamentally innovative it is a wave of concentrated product innovation that will happen thanks to TRansformers, LLMs, GPT-4 and multiple other trends that stack fairly well. It’s definately a Buff, if not a major level-up.
Over the weekend I wrote a massive article about Generative A.I. that goes out on my Machine Learning Times Newsletter, here is the sign-up. While I can appreciate Peter or Cathie Woods giving us good signals about the importance of A.I., Elon Musk promising us Tesla robots that are general purpose agents to service humanity doesn’t get me excited like it used to.
I think what 2020 best represents is a convergence of all these different things in digital transformation. Just as we have to look back and say a bunch of hyped things failed:
Crypto and blockchain
Chatbots and IoT
VR and AR
Cloud gaming and Advertising based networks
Smart Speakers (Alexa)
Silicon Valley’s leadership
If we do not admit the above failed, we cannot easily know what’s next. We need to evaluate tech hype on the back of failed hype and here’s a lot of it in recent memory.
People like Peter and Cathy belong to the past. They belong to a generation of promisers whose life span sadly will miss out on the real good stuff.
According to Zion Market Research, the global artificial intelligence industry is expected to grow from $59.7 billion in 2021 to $422.4 billion by 2028. Virtually every industry is being disrupted by AI, automation, and robotics.
We need to be cautiously optimistic and ultra-realistic. We may experience a new Venture Capital environment which is a drag to the cost of innovation that could offset the speed at which Generative A.I. impacts our actual lives. A.I. and reality will intersect in new ways, prompt-based software 3.0 is a trend that could have some unexpected advances.
We’ll know more as GPT-4 is announced.
All of A.I. is a Multi-Decade Paradigm Shift
I don’t see any evidence that it’s a paradigm shift, it’s just an upgrade to an A.I. trend that will take decades to even finish its first phase of business integration.
For consumers it’s most likely the BCI is the device that inherits the digital transformation of the A.I. paradigm shift, circa 2020 to 2060. Then you need functional Quantum computers of scale to upgrade the machine learning.
So many large VCs have nearly been specialized in funding the A.I. paradigm shift.
Softbank Group
Masayoshi Son who owns this Japanese MNC group is known to have devoted 97% of his ‘time and mind’ to AI development. Its technological development fund, Softbank Vision Fund is the largest in the world at $93 billion out of which $28 billion has gone into AI-focused projects.
Lightspeed Ventures
A US-based VC firm is into AI/ML investments for early-stage projects. Its tech portfolio includes companies that were later acquired by Oracle, Texas Instruments, Blackberry, Walmart, and other corporate biggies. Their biggest investment includes $ 7 million in People.ai, an AI company known for developing applications for revenue optimization.
Andreessen Horowitz
It is one of the most well-known names among venture capital firms investing in AI. With its humble beginnings in Silicon Valley, California in 2009, it has now reached the status of venture capital unicorn investing in all stages of company development. Apart from raising $30 million for People.ai, it has a dedicated bio fund to help medical AI companies.
Institutional Venture Partners
It is a US-based private equity fund, largely involved in providing funds to companies in later stages of development. IVP has become a major investor in AI startups with its investments in companies like GitHub, Soundcloud, and Indiegogo. The total investments in AI and ML startups add up to $7 billion.
Two Sigma Ventures
It is a subsidiary of Two Sigma Investments focussing on early-stage capital that focuses on IT and Computer technology projects aiming at bringing positive change in the world. Amper Music, an AI software for making music for videos based on emotional mapping, Zymergen, an AI software to improve biological manipulation, and Socore, a predictive analytics software are some of the AI companies its funds are parked in.
Toyota AI Ventures
The VC subsidiary of Toyota company is aimed at funding AI projects in robotics and autonomous mobility. With $100 million as seed capital, its emphasis lies largely on being able to provide expert advice and guidance from a team of experienced scientists and entrepreneurs. Some of the major projects include elementary robotics, robots for home care, etc.
Y Combinator
As a business incubator, they are experts in investing small amounts of money in large numbers of AI/ML startups. After investing $150K twice a year into startups, they focus on strategic planning for 3 months working closely with company experts. The projects it has so invested in include May Mobility, Lyre Bird, Msg.ai, etc.
Data Collective
Having funded 20 AI startups in the last 7 years, it holds the record for backing the most AI companies since 2012. Its main focus lies in funding Big Data companies operating at a global scale. Companies like Plenty, Recursion Pharmaceuticals, and Vicarious are in their portfolio.
Comet Labs
An incubator and fund it mainly invests in AI and robotic startups. They are also into partnering with industry experts to provide startups with an environment conducive to their growth, in technology and business realms. The projects which have received their funding include Creator, Iron Ox, Prenav, etc.
Accel
A veteran in VC funding is known for choosing companies with the potential for propelling into the next generation. Having joined the AI revolution recently, its investment portfolio includes $650K into AceBot, and projects such as Scale, SigTuple, and Facilio.
Now with the Generative A.I. and Software 3.0 Movement other Funds will also become more significant in A.I including:
Conviction
Coatue
Others (?)
Maybe once the Metaverse is complete circa 2030, a retired Cathie Wood and Peter can have tea. Venture Capital will have to focus on early-funding to get the best possible ROI, they will no longer be able to splash cash like they have for Stability.AI and Jasper.AI this year alone. Are $100 million bets on hype a good idea, really?
TL;DR so no Generative A.I. is not a new paradigm, it’s a new feature of the A.I. paradigm that honestly takes a lot of time to get anywhere interesting. Think about it, Amazon axed nearly its entire Alexa division.
Softbank and a16z bled money in their bets on A.I. and crypto most recently. This isn’t a game, this is us trying to build a more sustainable future for 8 billion people!
Thanks for reading!