AI predictions for 2024: What top VCs think

As 2023 draws to a close, it’s a time of reflection on the monumental advances — and ethical debates — surrounding artificial intelligence this past year. The launch of chatbots like Bing Chat and Google Bard showcased impressive natural language abilities, while generative AI models like DALL-E 3 and MidJourney V6 stunned with their creative image generation.

However, concerns were also raised about AI’s potential harms. The EU’s landmark AI Act sought to limit certain uses of the technology, and the Biden Administration issued guidelines on its development.

With rapid innovation expected to continue, many wonder: What’s next for AI? To find out, we surveyed leading venture capitalists investing in artificial intelligence startups for their boldest predictions on what 2024 may bring. Will we see another “AI winter” as hype meets reality? Or will new breakthroughs accelerate adoption across industries? How will policymakers and the public respond?

VCs from top firms including Bain Capital Ventures (BCV), Sapphire Ventures, Madrona, General Catalyst and more offered their outlook on topics ranging from the future of generative AI to GPU shortages, AI regulation, climate change applications, and more. While perspectives differ on risks and timelines, most agree: 2024 promises to be a defining year for artificial intelligence. Read on for their boldest predictions and insights on what’s to come in AI.

The rise and fall of generative AI startups

“Many generative AI companies will die. If you weren’t one of the startups that raised monster rounds this year, the future will be uncertain for you. Many [generative AI] companies will compete with each other, startups built on top of OpenAI will experience platform risk, and fundraising into these companies will dry out. At Day One, we’ve stopped looking at these deals altogether. 

I’m excited about AI in biotech, genome, climate and industrial applications. AI will save lives by helping scientists and researchers develop new treatments and diagnostics based on human genomic data. In the psychedelics industry, our portfolio company Mindstate is using AI to create new “states of mind” based on their largest comprehensive data set of trip reports to help treat treatment-resistant PTSD. AI in fertility, reproductive and longevity will completely alter the paradigm of humans’ lifespan and how we have children. In climate, companies are using AI to defend our ecosystems, like how Vibrant Planet is using AI/ML to prevent catastrophic wildfires that are happening globally.

AI will also be able to read what’s happening in people’s minds and project images of their thoughts—it’s quite fascinating and I’m curious to see how it’ll unlock knowledge about human consciousness and unconsciousness.

Masha Bucher, Founder and General Partner at Day One Ventures

Convergence of data modalities in multimodal models

“In 2024, the convergence of data modalities—text, images, audio—into multimodal models will redefine AI capabilities. Startups leveraging these models will enable better decision-making and improved user experiences, including personalization. We will see novel and transformative use cases across industries like manufacturing, e-commerce and healthcare. On the infrastructure side, AI workloads will become more demanding, and I expect to see innovation around multimodal databases. While not every use case will require multimodal models, first generation LLM startups in many sectors will face new competition and the pressure to continue to innovate and build defensibility will be intense.”

Cathy Gao, Partner at Sapphire Ventures

“We’re going to see multi-modal retrieval & multi-modal inference take center stage in AI products in 2024. AI products today are mostly textual. But users prefer more expressive software that meets them in every modality, from voice to video to audio to code and more. If we can get these architectures to work at scale, we could unlock software that provides much more accurate and human results, from drawing the answer to making calls in your tone and voice so you can attend less meetings to converging on the right outcome via collaboration with other AI and human entities. To power this, we expect ETL providers like Unstructured to diversify to include new data sources, more startups making use of the Segment Anything architecture from Meta, and startups like Contextual becoming full-scale solutions for multi-modal retrieval.” 

Rak Garg, Vice President at Bain Capital Ventures

“We continue seeing AI blossom into more use cases, especially in industries that are both large and rusty. In healthcare we are excited about the possibility of using computer vision to detect cancer, for using machine learning to help with better diagnoses, and of using generative AI to reduce paperwork. Consider we spend $4.3 trillion in healthcare in the US, which as a percent of our GDP is almost double of the average of OECD, and that almost a third is just in administrative costs… And while Chat has been a key buzzword of 2023, companies need to think beyond Chat in 2024.  Multimodal AI across input, training, model creation and output are key areas of innovation.”

Amit Garg, Managing Partner at Tau Ventures and Sanjay Rao, Managing Partner at Tau Ventures

“Multimodal models will make it much easier to create compelling interactions with AI agents and the quality of the AI will make it nearly impossible for humans to discern the difference between a computer and a human in certain use cases. We can see this already in places like Character.AI and Instagram and expect this to take hold in the workplace in areas like training, customer support, and marketing / sales. You will be building a relationship with a machine sooner than you know.”

Jess Leao, Partner at Decibel VC

Democratization of AI through open source

“We predict that more open-source models will be released in 2024, and we are especially looking at large tech companies to be one of the major contributors. Some examples of this could include companies like Tesla, Uber and Lyft (historically both big contributors to open-source projects), and even Snowflake. We would not be surprised if some of these models spun out into companies and received large funding rounds.”

Vivek Ramaswami, Partner at Madrona and Sabrina Wu, Investor at Madrona

“I see multimodal becoming the de facto standard for any large model provider by H2’24. The marquee model builders who have historically maintained proprietary models will begin open-sourcing select IP while releasing new benchmarks to reset the conversation around benchmarking for AGI.”

Chris Kauffman, Partner at General Catalyst

“There is a race-to-the-bottom in generative AI pricing between OpenAI, Mistral, Google, and others serving open-source models. Most are incurring losses using the existing hardware infrastructure (evidenced in per-token input/output costs) and hoping to make it up on volume. The imperative for generative AI companies is clear: find pathways to profitability and scalability. Based on this need, I believe VC investments will go toward developing efficient models, leveraging new AI compute hardware, and providing value-added services like industry-specific model fine-tuning and compliance.”

Jimmy Kan, Partner at Anzu Partners

GPU shortage: A persistent problem or a temporary setback?

“2024 will be the year of real time diffusion applications. In 2023, we saw some major theoretical improvements in diffusion model inference speeds — such as the original consistency models paper by Song et al, and, more recently, LCMs. (Also, Adversarial Diffusion Distillation.) We’re already starting to see projects that use these ideas, such as Dan Wood’s Art Spew (77 512×512 images per second, on a single 4090), Modal’s turbo.art (based on SDXL Turbo), and fal.ai’s 30fps face swap. In 2024, we’ll see more realtime image, audio, and video generation diffusion applications.” 

Slater Stich, Partner at Bain Capital Ventures

“The GPU shortage continues to ravage the startup ecosystem, making it hard for new companies to bring their products to market. There are two ways to solve this problem; either new compute options emerge that break free of the Nvidia monopoly on AI, or new models/architectures emerge that are more efficient with compute resources. I expect to see large amounts of funding go towards novel model architectures that run in linear, not quadratic time, such as Mamba from Cartesia AI, in addition to platforms built around diffusion models and liquid neural nets as a faster, cheaper, more performant alternative to transformer-based LLMs.” 

Rak Garg, Vice President at Bain Capital Ventures

“For starters, the GPU shortage is not necessarily as acute or definitive as everyone might think. The bigger issue is the famine of utilization for existing infrastructure therein, which I believe will persist in 2024 alongside continued supply chain constraints. Fixing lower-level software for AI will be the key to resolving the illusory GPU ‘shortage’”’ and much more real utilization issues. Until then, the only short-term solution we have is, simply, more computation. That said, I predict GPU constraints to persist in 2024, with the NVIDIAs of the world experiencing continued backlogs, while competitors (namely AMD and Intel) will each gain 1-2% of GPU market share as a result of demand-side desperation.”

Chris Kauffman, Partner at General Catalyst

“A contrarian take is we will eventually *not* have a GPU shortage. The market will converge to a small handful of buyers and suppliers. Nvidia and others will scale up to meet forecasted demand, and Microsoft, Google, Amazon, Facebook, and many sovereign nations will still be large buyers. The rest of us will rent them from the cloud providers but there will be plenty of leasable capacity to go around. The rich will get richer, but quality of life will improve for the ‘GPU-poor’.”

Jess Leao, Partner at Decibel VC

“According to the Taiwan Semiconductor Manufacturing Company (TSMC) Chairman, ‘it’s not the shortage of AI chips. It’s…our COWOS [advanced chip packaging] capacity,’ and advanced memory and packaging capacity will ramp; however, the long-term sustainability of AI in production won’t rely on general-purpose GPUs like Nvidia H100 and AMD MI300X. Investments will shift to focus on hardware specialized for inference, rather than training. NPU innovations like d-Matrix and EnCharge AI, utilizing near/in-memory computing, are emerging as cost-effective and environmentally friendly solutions, suitable for deployment both on local AI PCs and within data centers.”

Jimmy Kan, Partner at Anzu Partners

Apple and Google: Sleeping giants?

“We believe 2024 will bring some big releases from Apple, perhaps even their own GPT.  There have been reports of an Apple LLM known internally as Ajax GPT. While the model was created for internal use, next year we could see Apple making Ajax (or related models) more public, or incorporating generative AI capabilities across its apps (e.g. XCode, Apple Music) and devices (e.g. Mac, Vision Pro). And while this was more machine learning than AI, just last week Apple released MLX – an “efficient machine learning framework specifically designed for Apple silicon.” Releases from Apple could have massive influence over not only existing models but how the US approaches regulation, given Apple’s prominent role as a consumer device manufacturer.”

Vivek Ramaswami, Partner at Madrona and Sabrina Wu, Investor at Madrona

“If 2023 was the year of Open AI and Microsoft owned the airwaves, next year we will all be talking about Google. Google’s substantial investment in Gemini and unrivaled data and compute resources will offer developers GPT-4+ capabilities in all shapes and sizes, pushing the frontier for all foundation model providers. Don’t rule them out just yet.”

Jess Leao, Partner at Decibel VC