2024 Predictions and Looking Back at 2023’s

Posted on December 21, 2023

The most ambitious ideas have yet to be discovered. That conviction is what drives us day in and day out to strive to find and help the brightest founders take flight, the most powerful data science reach its potential, and the most impactful projects come to life.

As we look to 2024, here are our predictions for what innovations are going to shake up the world as we know it.

💡 If you’re a founder pursuing any of these – or other data science and advanced computing-driven ideas that deviate from the norm – we want to hear from you.

But before we do that, we wanted to take a look at our predictions from 2022 (see blog post here). Last year, we were excited and optimistic for technology to propel humanity forward and positively impact our lives and society. We also shared insight into AI startup opportunities they are most bullish about, and why (see blog post here). 

There are a bunch of things we got right last year. Our predictions blog post dropped on 11/29/22 and included Villi’s prediction that large language models and genAI would emerge as foundational technologies. The next day, OpenAI announced ChatGPT (we swear we had no heads up). Frances has written about how vertical search would grow (and we made an investment in that space), and Colin forecasted a resurgence in technology-driven North American manufacturing (where we made a 2023  investment). Time will tell if we were right about some of our other predictions including an improvement in the building blocks of web3, or our continued belief that we are experiencing a sea change in biology and life science enabled by AI and data. But looking back, we’re happy with our predictions from a year ago and most importantly how these theses led us to founders that we have had the privilege to back pursuing these exciting goals.*

So on to 2024…

10 Predictions for 2024

(1) Neuroscience is the next frontier – most ailments of the nervous system have no effective therapies. The next generation of startups will use computing and data science to change the status quo – Dusan Perovic

These diseases are highly heterogeneous. Almost every Alzheimer’s patient is unique – patients progress at different speeds and their diseases are likely driven by different factors. To begin to understand this complexity we need to put together a patient’s genomic, clinical, and imagery data. With the ML tools we have now, we can finally do this,  which I predict will lead to a better understanding of these diseases and a faster path of novel targets that we can go after. In addition, you have probably read about the success of obesity drugs. These drugs act on the brain circuit that controls if/when we feel hungry. The success has been so strong that the market cap of Novo Nordisk, the Ozempic manufacturer, earlier this year surpassed the GDP of Denmark, where the company is headquartered. I expect this market tailwind to lead to more interest in neuroscience, and a lot of new startups to be built in the coming years.

(2) Generative Models will become foundational in business applications – Villi Iltchev

In 2023 we saw the rapid rise and evolution of Generative AI, which fueled tremendous adoption of these powerful models in a wide range of business applications. We believe this is just the beginning of a large investment cycle that will disrupt, reinvent, and reimagine business applications. In addition to what I see as a natural opportunity to disrupt incumbent software vendors with modern intelligent applications, I am particularly excited to see Gen AI startups in the following categories:

  • Applications with dynamically generating UI. I think founders may rethink how we interact with software applications entirely. I can envision a future where applications have dynamic UI that changes based on context, user, and task.
  • Surface Area. Gen AI may enable founders to bundle point applications into integrated solutions and bridge what were previously separate software categories and applications.
  • Business Model. I expect Gen AI will create an opportunity for significant business model disruption in many sectors. We may see the rise of startups that turn industries upside down by charging for a unit of output/results vs. a unit of time, for example. 
  • Language as Code. Gen AI can empower non-technical users to build and interact with complex systems, perform analytics, and build software. New applications that provide superpowers to non-technical users will likely emerge.

(3) Creatives will soon use AI-driven platforms to create complex works of art such as films and video games at 99% lower cost – Dan Abelon

Individuals will be able to use software to create world-class entertainment that would have previously taken months and millions of dollars. These software platforms will be enabled by a combination of (1) Specialized AI models for each creative element – characters, video, dialogue, music, etc, (2) Delightful workflows to stitch the pieces together, and (3) Founding teams who have both extensive creative experience as well as AI expertise.

(4) LLMs will unlock a once-in-a-generation transformation of the legal industry – Frances Schwiep

Conventional wisdom regarding AI’s market impact has been flipped on its head. Until 2023, the prevailing view was that AI would first take over low-skilled and repetitive tasks, particularly in back-office settings. Yet, thanks to OpenAI and LLMs, AI is showing potential to be most disruptive in high-skilled, deep-knowledge-based professions. 

In the medical field, we saw GPT/ AI-based models passing the MCAT (in some cases surpassing the average pass rates by a long shot) and profound impacts in drug discovery, gene sequencing, and much more. In a similar vein, I believe LLMs and AI are on the cusp of transforming the legal industry.

Law is particularly well-suited for LLM disruption, given the profession is inherently a language-based practice, but with formulaic, logic-based elements. The legal industry is a $400 billion market in the US alone, and even small efficiency gains can unlock a significant amount of value. How does this happen? I believe it includes:

  1. Accelerating the speed and accuracy of legal analysis, 
  2. Creating opportunities to reduce friction by generating standard legal documents on the fly
  3. Enabling entirely novel workflows and 
  4. New business models emerge such as value-based billing.

(5) Despite expected setbacks, Autonomy (not just cars, but many types of machines) will cross chasms and start becoming more mainstream – Colin Beirne

At our annual meeting in September 2023, I predicted that the majority of the people in the room would, by the end of 2024, have taken a ride in an autonomous car. A few weeks later, one of two major robotaxi services in the U.S. (temporarily) shut down its service after an incident involving a pedestrian and another car driven by a person, leaving just one operator of autonomous taxis. Nevertheless, I think after decades of work, we’re about to turn a corner in autonomy. Cars are the most visible instance of this, and we’ll see if my prediction proves correct. But in many other sectors – heavy equipment, security, agriculture, industrial manufacturing, and more  –  autonomy is coming quickly, helping to offset skilled labor shortages and drive better unit economics. While we may not yet have our personal home robot butler, our lives will increasingly be enabled by autonomous machines working behind the scenes. 

(6) We will see a resurgence in the crypto industry driven by founders building products and services targeting mass market, real-world use cases – Andy Kangpan

Despite the overwhelming market consensus that crypto is dead, the industry has charged ahead in 2023 on several dimensions. From an infrastructure perspective, the industry is entering its broadband era with core technologies being orders of magnitude more scalable and reliable than even a year ago. Additionally, major financial institutions and corporations (e.g., Societe General, Blackrock, Goldman Sachs, etc.) are continuing to throw their hat in the ring pursuing use cases from real-world asset tokenization to payments infrastructure. Taking these trends together, I believe 2024 will be a hallmark year for the industry characterized by a new wave of founders building businesses built on real-world use cases that touch a much wider range of users than we’ve seen before.

(7) 2024 will bring an unprecedented number of partnerships between cutting-edge technology companies and traditional financial institutionsKyra Durko

FIs will struggle to hire top talent to build novel, AI-first products. This is great news for startups that have application or infrastructure layer solutions to offer to banks. It will become easier to start conversations with executive teams at these institutions throughout 2024. I predict FI’s next generation of customers will expect their banks and financial tools to capitalize on new technologies and will flock to providers that execute this roadmap best. FIs’ ability to move fast and execute this strategy will be crucial to remaining relevant in years to come.

(8) New deep learning platforms will fundamentally transform the scientific discovery processGabriella Garcia

New deep-learning approaches will dramatically change how we approach scientific discovery, going beyond traditional methods in fields like math proofs, protein biology, climate modeling, chemistry simulations, and material science. In this new shift, deep learning uses simulations/emulators based on scientific equations, not just empirical data, making forecasting and experimentation more accurate, scalable, and efficient, and can even leverage cross-domain knowledge by analyzing diverse data modalities. It’s playing a crucial role in areas like forecasting the weather, generating new metal alloy compositions, optimizing the design of fusion reactors, or calculating the binding affinities of candidate drug molecules to a target protein – opening up new possibilities for treatments and sustainable technologies.  Data science-driven startups have a first-mover advantage to commercialize their platforms, don’t have the tech debt of previous approaches, and I believe will be key in pushing forward this ‘fifth paradigm’ of scientific computing by hiring cutting-edge talent coming out of academic labs and research institutions. I’m bullish on “AI4Science” and startups at the forefront of exploring these vast opportunities and how they’re set to change how we understand and explore the natural world.

(9) AI-first ETL tools will accelerate enterprise LLM usage on proprietary data  – Vin Sachidananda

In the past year, we have seen pervasive gaps in enabling enterprises to ingest and transform data, across various formats, in a manner best suited for operation with/by LLMs. This gap is predicated on the differentiation of these data pipelines compared to those within Web/ML workflows. In particular, AI data pipelines look very different, in terms of input/output stream, from previous ones which transformed data from one structured format to another. There are opportunities to define best practices for unstructured inputs/vector outputs in lineage, drift, versioning, etc. Additionally, the multimodal capabilities of Generative AI models have strong ramifications in extending what modalities can be indexed together in a data store. Lastly, the use of LLMs both in the indexing and generation within data stores leads to several challenges and opportunities. A range of composable transforms based on varying prompts, chain of thought reasoning, and across a range of models provides new capabilities but also requires strong, extensible tooling to produce/evaluate the best outputs in an ETL pipeline. We believe that new offerings will emerge, both from cloud service providers and nascent startups, to bridge this gap and enable any data to be operable/transformable for use with LLMs.

If you’re a founder building a data science-enabled business, we want to hear from you: platform@twosigmaventures.com

*View all TSV investments here.

The views expressed herein are solely the views of the author(s), are as of the date they were originally posted, and are not necessarily the views of Two Sigma Ventures, LP or any of its affiliates. They are not intended to provide, and should not be relied upon for, legal, regulatory and/or investment advice, nor is any information herein any offer to buy or sell any security or intended as the basis for the purchase or sale of any investment.  The information herein has not been and will not be updated or otherwise revised to reflect information that subsequently becomes available, or circumstances existing or changes occurring after the date of preparation.  Certain information contained herein is based on published and unpublished sources. The information has not been independently verified by TSV or its representatives, and the accuracy or completeness of such information is not guaranteed. Your linking to or use of any third-party websites is at your own risk. Two Sigma Ventures disclaims any responsibility for products or services offered and/or information contained on any third-party websites.