Has innovation taken a backseat in the Generative AI landscape, or is the financial muscle now the sole architect of a successful LLM? Let's dive into the big changes and trends we saw in the first quarter of 2024 and think about what's really driving the industry forward.
In January, we shared an article “Cutting Through the Noise in the Generative AI Landscape as an Early-Stage Investor” where we discussed the rapidly evolving Gen AI market and presented our investment hypotheses for the sector. Here, we used a framework inspired by A16z to analyze the AI market, identifying key players and their roles, which helped us understand the commercialization and integration of AI technologies in products and services.
Initially, our thesis identified two areas within the Gen AI tech-stack as central to our focus when looking for companies to invest in.
Infrastructure software providers are essential for GenAI, offering critical resources for AI model development and training. On the other hand, end-to-end application companies develop comprehensive AI solutions, often with proprietary models, earning the "end-to-end" label by controlling the entire AI application process from data input to final output.
Since January, there have been significant developments in this space. With this in mind, our aim is to provide a summary of the investment dynamics that unfolded in Q1 2024. We will focus on notable trends and emerging companies within the northern region. Therefore, this article will cover the following:
Reflecting on Q1 2024, it's clear that Generative AI has continued to captivate the investment community, building on the significant momentum established in 2023. According to a Synaptic report, AI startups managed to secure $42.5 billion through 2,500 equity rounds in 2023. This has also continued into the early months of 2024, with January and February alone witnessing 250 investment deals in the AI sector, setting the stage for further growth.
One of the notable events of this quarter occurred on March 20, when Astera Labs, an AI cloud company backed by the likes of Sutter Hill Ventures and Fidelity, made its debut on the public market. This move, as reported by Reuters, reflects a broader market trend of increasing IPOs and M&A activity within the AI domain. Commentary from CFO news state that “if these deals don’t dazzle, we should not expect to see investment bankers and late-stage startups throw in the towel in 2024”. Similarly, TechCrunch hinted that the success of Astera Labs' IPO will reveal how much investors want in on AI.
In another notable development, Nvidia's stock has experienced a remarkable rise, soaring by 82% in 2024 after a previous year's jump of 239%. This performance draws parallels to the dot-com era's giants. Jess Summer at New York Times stated that ”if NVIDIA keeps rising like this, it will be bigger than the global economy”.
As the quarter came to a close, it has been evident that more acquisitions in the AI sphere have been taking place. Apple made a significant move by acquiring Darwin AI in March, shortly followed by Bain & Company's strategic acquisition of PiperLab. Not to be outdone, Amazon announced a $4 billion investment in Anthropic, an AI safety and research firm, in early April, highlighting the tech giant's commitment to the forefront of AI innovation.
The landscape of AI applications can be broadly categorized into two types: vertical and horizontal. Vertical AI applications, as noted by Forbes, are bespoke solutions crafted to address the distinct needs of specific industries. This specialization allows them to harness industry-specific data and insights to solve particular challenges. The value of vertical AI lies in its depth—integrating domain expertise into its algorithms, it delivers precisely targeted solutions that can resonate with the industry's core operations.
On the flip side, horizontal AI applications offer a broad spectrum of services that are applicable across various industries and use cases. These solutions are versatile and 'cross-functional', with the ability to address common challenges that are not unique to any single industry.
This difference is significant because it shows a pivot in the AI sector towards more niche, specialized solutions that can provide deeper value in particular areas, as opposed to the one-size-fits-all approach that horizontal solutions might offer. With vertical AI, industries can expect more tailored, knowledgeable, thus potentially more effective AI-driven improvements in their specific contexts.
As investors we recognize the merits of both vertical and horizontal AI applications. However, our conviction is stronger in vertical applications, as they possess a more robust long-term technical advantage. They are often underpinned by proprietary models and require the accumulation of substantial high-quality data, a process that is both time-intensive and access-dependent, making these solutions less vulnerable to quick replication. This “data barrier” contributes to their defensibility and positions them favorably for sustained success.
To showcase recent progress in the GenAI space, we’ve mapped out notable investments rounds from the past nine months (Aug’23 – Apr’24) grouping companies into two primary layers: infrastructure and application.
In the application layer, we've distinguished between vertically focused companies that serve specific industries and horizontally focused ones that cater to multiple sectors. This mapping serves to highlight where investments are being channeled and provides insight into the evolving landscape of GenAI technology and applications.
Hailo.ai (raised Apr’24): Breakthrough AI processors for deep learning applications on edge devices. Israel, founded in 2024, Infrastructure Layer. Raised $344 million in early-stage funding from SKC, Capricorn Investment Group.
Vizcom.ai (raised Mar’24): AI for industrial product visualization and design. US, founded 2021, Infrastructure Layer. Raised $25 million in early-stage funding from Index Ventures, Unusual Ventures.
Together.ai (raised Mar’24): Open-source AI development platform. US, founded 2023, Infrastructure Layer. Raised $228 million in early-stage funding from NVIDIA, Salesforce Ventures.
Anthropic.com (raised Feb’24): Building chatbot Claude. US, founded 2021, Infrastructure layer. Raised 7.6 billion since inception from Google, Amazon, Salesforce, and Menlo Ventures.
Photoroom.com (raised Feb’24): AI-enabled app for studio-quality product photography. Paris, founded 2024, Application Layer, Vertical. Raised $62.1 million in early-stage funding from Aglaé Ventures, Y Combinator, Balderton Capital.
ElevenLabs.io (raised Jan’24): Voice AI tools for multilingual audio content creation. US, founded 2022, Application Layer, Horizontal/Cross-functional. Raised $101 million as Unicorn from Disney Accelerator, Andreessen Horowitz.
Qdrant.tech (raised Jan’24): Open-source vector search engine/database for AI applications. Germany, founded 2024, Infrastructure Layer. Raised $28 million in early-stage funding from 42 CAP, Spark Capital.
X.ai (started talks about raising): AI chatbot providing real-time information via social media. US, founded 2023, Infrastructure Layer. Raised $135 million as Unicorn from DCM Ventures, FirstMark Capital. Started talks about raising $3 billion as of Apr’24 according to Fortune.
Mistral.ai (raised Dec’23): Open and portable GenAI solutions for developers and businesses. France, founded 2023, Infrastructure Layer. Raised $544 million as Unicorn from Databricks Ventures, Andreessen Horowitz +.
Inflection.ai (raised Jun’23): Personal AI chatbots. US, founded 2022, Infrastructure Layer. Raised $1.5 billion, classified as Unicorn, with investors like Bill Gates, Reid Hoffman, NVIDIA.
Modal.com (raised Apr’24): Cloud-based ML infrastructure. US, founded 2023, Infrastructure Layer. Raised $32 million in early-stage funding from Redpoint, Left Land Capital.
Character.ai (raised Mar’24): Conversational applications enabling creation and interaction with AI characters. US, founded 2021, Application Layer, Horizontal/Cross-functional. Raised $150 million as Unicorn from Andreessen Horowitz, SV Angel.
Vizcom.ai (raised Mar’24): AI for industrial product visualization and design. US, founded 2021, Infrastructure Layer. Raised $25 million in early-stage funding from Index Ventures, Unusual Ventures.
Leya.law (raised Mar’24): AI assistants for legal professionals to answer complex queries and optimize workflows, part of YC2023. Sweden, founded 2024, Application Layer, Vertical. Raised undisclosed amount in early-stage funding from Benchmark Capital.
Flock.io (raised Mar’24): Decentralized AI for privacy preservation. UK, founded 2022, Infrastructure Layer. Raised $8 million in seed-stage funding from Lightspeed Faction and Tagus Capital.
Buddywise.co (raised Mar’24): Computer vision and ML for industrial safety monitoring. Sweden, founded 2020, Application Layer, Vertical. Raised $5.1 million in seed-stage funding from Kvanted, J12 Ventures, Norrsken VC.
Perplexity.ai (raised Jan’24): AI-powered generative search engine. US, founded 2024, Infrastructure Layer. Raised $102 million in seed-stage funding from Elad Gil, IVP.
Olakrutrim.com (raised Jan’24): Specializes in AI compute chip creation and cloud infrastructure. India, founded 2023, Infrastructure Layer. Raised $73 million in seed-stage funding from Matrix Partners.
Raga.ai (raised Jan’24): AI testing platform for secure and reliable enterprise models. US, founded 2022, Infrastructure Layer. Raised $4.7 million in seed-stage funding from TenOneTen Ventures, Mana Ventures.
En.Tacto.ai (raised Dec’23): Optimizes procurement for compliance and efficient material sourcing. Germany, founded 2020, Application Layer, Vertical. Raised $56 million in early-stage funding from Sequoia, Index.
Armada.ai (raised Dec’23): Full-stack edge computing platform. US, founded 2022, Infrastructure Layer. Raised $60.3 million in early-stage funding from Shield Capital, Lux Capital.
EnChargeAI.com (raised Dec’23): Advanced hardware and software for AI in edge computing. US, founded 2022, Infrastructure Layer. Raised $44.3 million in early-stage funding from RTX Ventures, Alumni Ventures +.
Lynxtech.com (raised Nov’23): AI solutions for financial crime and fraud detection. UK, founded 2023, Application Layer, Vertical. Raised $18.6 million in early-stage funding from Santander Bank, Forgepoint Capital.
Runpod.io (raised Nov’23): GPU compute provision for ML developers. US, founded 2022, Infrastructure Layer. Raised $2 million in seed-stage funding.
Cranium.ai (raised Oct’23): Security and trust platform for AI models. US, founded 2022, Infrastructure Layer. Raised $36.7 million in early-stage funding from Telstra Ventures, KPMG LLPs +.
Poolside.ai (raised Aug’23): AI platform to improve software development efficiency. France, founded 2023, Infrastructure Layer. Raised $126 million as Unicorn from Xavier Niel, VC Felicis.
AilyLabs.com (raised Aug’23): Analytics service platform for personalized business insights. Spain, founded 2020, Application Layer, Horizontal/Cross-functional. Raised $20.5 million in early-stage funding from Insight Partners.
Scriptme.io (bootstrapped): Rapid audio and video transcription using AI. Sweden, Application Layer, Horizontal/Cross-functional.
Multion.ai (bootstrapped): Personal agents to streamline user tasks and improve efficiency. US, founded 2023, Infrastructure Layer.
Concluding on this updated outlook in the Generative AI sphere, it’s evident that the landscape is not just growing – it’s exploding! Especially within the AI infrastructure layer. This includes pivotal areas such as model training, foundational models, and DataOps and Interface, highlighting a robust financial commitment to developing the AI backbone.
Despite the allure of foundational model companies capturing substantial funding and becoming unicorns, it's crucial to question the long-term viability and return on these massive investments. The real challenge lies not just in securing funding but in sustaining growth in an increasingly crowded space, making agility and innovation key differentiators.
In contrast, the vertical application sectors like legal, and security & compliance present a landscape of minimal competition, suggesting a ripe opportunity for groundbreaking solutions and investments. The focus of VC investments on startups catering to data center needs and AI chip development further underscores the industry's direction towards supporting intensive AI workloads.
Looking back at the first quarter of 2024 in the Generative AI space, with all mega rounds, we have to wonder: Are we overly fixated on funding, at the expense of genuine breakthroughs? It's crucial to figure out if real creativity is still at the heart of Gen AI or if it's becoming all about financial power. As we move into the next quarter, this question becomes more important, and it might just define where we're headed next in the world of Gen AI.
We continue to closely monitor developments in the AI space; stay tuned for our upcoming report after Q2 2024.
Read also
Cutting Through the Noise in the Generative AI Landscape as an Early-Stage Investor
AI Will Change the World More Than the Internet Ever Did
