The Buzz of Generative A.I. Heading into 2023
ChatGPT, Lensa A.I. and new A.I. labs and startups show the A.I.adoption wave of 2020s is real.
Hey Guys,
Apart from the commercial aspect of Generative A.I., on the back of BigTech and Venture Capitalists, where ChatGPT is taking to public relations extremes of virality, there is a certain relentless joy in innovation and trying to be a part of the moment.
The start of 2023 is a fairly good time to get involved personally if you wanted to start your own data engineering, Generative A.I. or AI trends Newsletter, I recommend you to to do it (you never know what it might lead to, especially if you live in India or another youthful demographic like Nigeria, Indonesia, Brazil, etc…)
Why Start a Newsletter in A.I., Data science or Tech?
If you wanted to be a digital nomad as a diversified solo-entrepreneur you could:
Substack subscriptions
Talent communities (pallet)
Sponsor Ads (Paved, Swapstack, others)
How it all can compound after a few months/years is worth exploring.
Where there is a will, there is a way. You could literally work in Spain, or all of these other places.
A.I Adoption Trends
We have to assume that 2023 will be even more formidable than 2022 in terms of A.I. adoption (poll). AI innovations continue to deliver big benefits to business and adoption rates will accelerate in coming years.
Generative AI is well on the way to becoming not just faster and cheaper, but better in some cases than what humans create by hand. How can this be? Up till just a few years ago we thought we were special in our ability to maybe write poetry, design products, make games and crank out code? It turns out GPT-4 and future iterations will do this better and better.
Venture Capitalism have significant incentives to steamroll A.I. adoption hype further and bigger than ever before in 2023. The reasons are simple, it’s highly profitable. According to data from PitchBook, investment in generative AI has considerably increased since 2020. Their data shows a 425% jump from 2020 to now, with $2.1B invested in 2022 alone.
We are entering an era of Artificial Imagination, and while it’s not AGI, buzzy startups like OpenAI have internal valuations of up to $20 Billion. It’s pretty incredible what Stability.AI and others will become.
Diffusion, Transformers, Large Language Models, hallucatinting human level A.I (HLAI) really makes one wonder and the saga of Generative Pre-trained Transformer models as so-called foundational models.
According to data from PitchBook, investment in generative AI has considerably increased since 2020. Their data shows a 425% jump from 2020 to now, with $2.1B invested in 2022 alone.
The VC Hype Cycle Dogma
People really do follow the money.
The Global Generative AI Market Size accounted for USD 7.9 Billion in 2021 and is projected to occupy a market size of USD 110.8 Billion by 2030 growing at a CAGR of 34.3% from 2022 to 2030.
Prisma Labs the maker of Lensa.AI have had considerable sales since their app went viral. It’s totally a Russian startup if you can believe.
The thing is while in 2022 we saw Stable Diffusion (by Stability A.I.), Midjourney, ChatGPT and Lensa A.I., in 2023 we might have a lot more winners. Big ticket winners too.
In the case of Prisma Labs, it co-founded by Oleg Poyaganov, Ilya Frolov, Andrey Usoltsev and Alexey Moiseenkov in 2016, and it evolved around its flagship app, Prisma. In 2018, Moiseenkov departed from Prisma Labs in favor of pursuing other ventures. After he resigned as CEO, Usoltsev took up the mantle.
As of early December, 2022 according to app analytics firm SensorTower, Lensa AI is also quickly climbing the “Top Grossing” app ranking on the App Store, meaning it is creating more revenue through app sales or (in this case) in-app purchases than almost any other app at the moment. I don’t have to tell you how much revenue that is? It’s $Billions. In a short period it likely made $Billions each day.
Lensa AI realized Generative A.I. + sex sells even better. Their “Magic Avatar” basically degraded women for profit. But all is fair in business, apps and A.I., I guess? So did women pay to get profile pictures that were more hot than their real-life pictures? I guess the verdict would be yes by the sales figures, and it all makes me wonder.
This combined with how good ChatGPT hallucinates believable content and it’s a bit frightening to see what will become of Generative A.I. in 2023 and the 2020s moving forwards.
Sam Altman must suddenly feel like his shadow A.I. lab turned sold-out for profit startup, OpenAI, is suddenly at the center of the world.
Real AGI there buddy. I’m not sure how companies like OpenAI or Sam Altman can even believe what they are saying at this point Generative A.I. has become such a game of Venture Capital maneuvering and gimmicks. Major disclaimers please:
OpenAI’s GPT-3 stands out: the model’s performance is a giant leap over GPT-2. The progress of GPT-3 to GPT-3.5 and davinci-003 and chatGPT (trained with RLHF) also goes to show what GPT-4 can become and what tools can be built upon it. I’ve found many don’t fully understand the leap from GPT-3 to GPT-3.5 which was after 2+ years of development.
The general narrative VC is pushing is that the content that generative AI can create is truly limitless and could be hugely transformative across numerous industries. This of course is false, but there will be many interesting tools that result. How useful or digitally transformative they are to actual work and creativity, of tasks and workflows, is the problem. However Microsoft seems to think GitHub Copilot for coders is worth selling to Businesses at a steeper rate. But more subscription revenue for Microsoft, is not to say that Generative A.I. adoption is great.
Hilariously text-to-images and ChatGPT is also causing real harm, chaos and content moderation problems. It’s not just Lensa A.I. sexualizing women or even portraying them nude, it’s what it does to real communities like DeviantArt, Stack Overflow, ArtStation, etc… we will find out more in 2023.
The hype leads to some unfortunate feedback loops and real people are duped into thinking this is real innovation. But you can cherry pick progress anywhere: New techniques, like diffusion models, shrink down the costs required to train and run inference. The research community continues to develop better algorithms and larger models. Developer access expands from closed beta to open beta, or in some cases, open source.
People want to believe an optimistic narrative, because it allows them to keep doing what they do with the incentives already in place.
Silicon Valley wants you to know that it’s still relevant, somehow even in a dystopia of Monopoly capitalism and winner-takes-all dynamics everyone from Canva, to Adobe to dozens of other big brands are scrambling to embed Generative A.I. and LLMs into their products.
What will be the result in 2023? With such an uncertain year for the global economy, “more hype” is the easiest answer.