Today we are diving into the annual CB Insights review of technology trends that are gaining momentum. This time there were as many as 20 such predictions (which I combined into 19), and more than 120 slides. Many are obvious and some are not. Let's see what investors should look out for and avoid. We've read the whole thing and picked the key points for you.
GPU shortage opens up new opportunities for startups
The excitement around artificial intelligence has led to a GPU shortage that rival manufacturers Nvidia and AMD simply can't keep up with. Elon Musk even joked about it at a recent conference, saying it is now harder to get a graphics card than drugs. Designing chips to work with AI remains a complex and time-consuming task. First, only a few facilities like TSMC can produce them. Second, only a few companies in the world can turn chips into an end product — an AI accelerator.
For a long time, graphics cards have been the domain of gamers, enthusiasts, and cryptocurrency miners, and haven’t caught big fish's attention. Now, however, GPUs have become the number one target of the world's leading corporations. Limited supply has forced entire governments to rush investments into local semiconductor manufacturing, but all this will only bear fruit in the distant future. So what to do now?
Look for startups that are developing tools to optimize GPU performance. By improving the accelerators’ efficiency, companies can save money by buying less hardware. What's more, startups can help existing chips run faster.
Startups to watch: CentML, ThirdAI, Run:AI, OmniML, Deci, CelestialAI, Graphcore, and Innatera.
The next stage in AI development is Multimodality
Multimodality means being able to retrieve and analyze information from different types of sources at the same time: for example, text plus photos and video. At the moment, AI is mostly limited to one modality, but its advancement is only a matter of time.
Startups can make good money by developing niche multimodal LLMs for different industries. Although AI is on the rise, its integration into business processes and operations remains a challenge. With advanced LLM models that are domain-specific, companies can gain an edge in the market.
Startups to watch: Artera and Ghost Autonomy.
Synthetic Data
Just as humans need books and other sources of information, AI requires data for learning and development. Publicly available datasets are running out, which will lead to a significant slowdown (if not halt) in the development of new LLMs.
According to analysts, we will no longer have quality textual data to train AI by 2026. To make matters worse, proprietary (i.e., private) sources remain inaccessible for this purpose, as using them is essentially theft. Synthetic data must come to the rescue.
The synthetic approach allows similar data to be developed from real one and in large quantities. Its quality and validity are sufficient for it to be used by giants such as Microsoft, Stability AI, OpenAI, and DeepMind. Companies using AI in their work are willing to pay a good price for synthetic data for their LLMs.
Startups to watch: Aindo, Scale, and Gretel.
DIY Software Development
Video cards are not the only scarce product on the market. Experienced programmers are also in short supply. Companies around the world are experiencing a labor crunch, which is why tools that make programming easier are gaining popularity.
You've probably heard of low/no-code platforms, but the latest trend is AI assistants, also called copilots. These tools allow companies to do programming in-house without having to hire outside experts. Saving resources and increasing staff efficiency make startups' products more in demand than ever.
Startups to watch: OutSystems, Retool, and Appsmith.
Human Resources: it could always be worse
In 2021 and 2022, HR startups suddenly received a huge surge of investment — over $2 billion. At the time, businesses were concerned about finding and hiring remote workers. This was for a variety of reasons, from a pandemic to a lack of local talent. However, it looks like God has stopped smiling at HR startups, which are going to have a tough time.
The top ten startups in this category attracted zero (!) investment in 2023. This is due to massive layoffs and downsizing in the tech industry around the world. In general, think twice before investing in such companies.
Quantum Computing is on the Horizon
Quantum computing is slowly but surely moving from science fiction to reality. Investment in startups has grown for three years in a row. The reason? It seems that quantum computing commercialization is just around the corner.
Neutral atom quantum technology recently made a real breakthrough, bringing the devices much closer to being used in the real world rather than in laboratories. Businesses are already rubbing their hands together as they prepare to take advantage of this new force in the market. In addition, quantum computing will give a strong boost to AI, allowing the development of new LLMs as quickly as possible.
The chances of getting into the captable of such a company are mediocre, but still higher than those of fusion startups with minimum cheques of hundreds of millions of dollars.
Startups to watch: Pasqal, NanoQT, QuEra, ColdQuanta, Planqc, and Atom Computing.
You can’t go wrong with Cybersecurity
Cybersecurity is one of the key areas of focus for the coming years. As businesses become more digitally driven, the risk of falling victim to cybercriminals increases. According to analysts, the damage caused by such incidents and various types of data breaches could range from $2 million to $100 million per attack. In addition, the number of cyber-attacks doubled last year compared to 2022.
The most important trend within cybersecurity itself is AI. Used not only for defense but also for attack, the future of cybersecurity will be a battle between one robot and another. The better AI wins.
Today, companies are more concerned than ever about their data security, so all doors are open for startups. Investment activity is particularly high here, from early to late stage. Keep an eye on this industry and don't miss out on these projects, they are very promising.
Startups to watch: Oleria, Gem Security, Silk Security, Aembit, Tessian, Armorblox, and Avanan.
Banks want AI and fear it at the same time
Artificial intelligence will have an impact on many sectors of the economy in one way or another, and banking is no exception. The industry recognizes the value of AI but remains too conservative and slow-moving to embrace such an innovation.
Bankers' biggest fears:
Data issues: privacy, sovereignty, and disparate locations;
Recruiting and retaining AI experts and data scientists;
Lack of budget;
Sufficient data sizes for model training and accuracy.
Bankers' top wishes for AI:
Financial document search and synthesis;
Enhanced virtual assistants;
Capital markets research;
Regulatory code change consultant;
Personalized financial recommendations.
AI startups will have to work hard to convince banks to become customers. No matter how good the product is, the main stumbling blocks are security, easy integration, and bank nature, whatever this means.
Startups to watch: Kasistom KoreAI and Aisera.
Blockchain: not even close
Blockchain has never been a star in the sky. In 2021-2022, the industry attracted unusually high investments of $27 and $29 billion respectively (compared to just $3.5 billion in 2020). But after that spike, funding plummeted to $8 billion in 2023. Low traction and minimal commercialization results are to blame.
Blockchain products remain on the fringes of attention, limited to isolated experiments. The technology has been around for years but has yet to prove its usefulness.
The banks are also unable to embrace the concept. Over the past two years, for example, several major blockchain trade finance networks have shut down, including WeTrade, Marco Polo Network, and Contour. The recent rise of Bitcoin is not saving blockchain either.
However, businesses don't hate blockchain as much as it might seem. It's just not a priority for companies, putting it at the bottom of the list. Look at other industries better.
Rising Natural Disasters boost Insurtech
Climate change continues to harm the global economy. In the US alone, a record number of natural disasters occurred in 2023, each causing more than $1 billion in damage. The phones of insurance companies are ringing off the hook, but statistics show that less than half of all claims are accepted. Some insurance companies have gone out of business as a result (StateFarm, Allstate, and Farmers Insurance).
All of this has forced insurance companies to invest in solutions that help predict and manage weather risk. There aren't many startups like this in the world, but if you come across one, don't miss it. The level of investment is relatively modest but stable and, most importantly, there is a real demand for the products.
Startups to watch: Pano, Mitiga, and Reask.
AI in the Pharmaceutical Industry
Artificial intelligence has found its way into pharma, and no wonder. The industry is particularly hungry for the computational power needed to research new drugs. Startups offering tools for developing small-molecule drugs (their non-small-molecule competitors are not as successful) are particularly popular.
AI not only makes the process faster but also more accurate, helping pharmaceutical companies bring new drugs to market much faster. And speed and accuracy cost money. Finding such a startup is no easy task, but you need to be sure of the deal's value. This is where investment activity remains a safe bet.
Startups to watch: Deepcurem Atomwise, Benevolent, Hexagon Bio, and Atomic AI.
Digital therapeutics (DTx) & wellness startups may need therapeutics themselves
The once-promising remote therapy industry seems to have run out of steam. Based on 2023 results, companies have raised just $0.7 billion, which is an anti-record for the past five years. Reasons for the failure include problems with insurance claims and the small number of customers willing to pay. We recommend ignoring these ventures for a time.
Startups are getting inside our brains. Literally
You've probably heard of Elon Musk's Neuralink, a company that experiments with the human brain using digital implants. Turns out it's not as much of a fad as it might seem. Preliminary results of experiments to stimulate certain parts of the brain are encouraging.
Investors believe the first implants could be used commercially for patients with mobility disorders. Clinical trials are already underway.
In addition, the implants have learned to read information from the human brain in conjunction with AI. The specific application of this technology is still unclear but sounds intriguing.
Startups to watch: Synchron, Paradromics, and Precision.
Robo Salesmen
The e-commerce industry continues to grow year-on-year, and it doesn't look like anything is going to stop it. Today, one of the main local challenges is the cost of acquiring a new customer, which is also increasing. For example, in 2013 it was $9 and in 2022 it's already $29 (a 222% jump). AI must come to the rescue.
Startups are offering retailers virtual sales agents in the form of chatbots. These can not only answer mundane questions but even haggle, upsell, and fully negotiate a deal. In general, they behave (almost) like a real manager. Virtual salespeople are proving particularly popular in some countries, such as India.
Besides, the introduction of digital assistants increases sales efficiency and reduces staff costs.
Startups to watch: Nibble, Zowie, Batna, Ergo, Tidio, Manychat, Loyal Guru, and Session AI.
Robo Cops
E-commerce suffers from another challenge: inventory loss due to errors and theft, or shrinkage. In the US alone, retailers estimate that shrinkage costs them $100 billion a year. No wonder, companies are looking for ways to secure their goods. Are they turning to AI again? Yes.
In 2023, loss prevention tech companies are raising massive rounds of funding. Startups offer a range of solutions: radio frequency identification (RFID), real-time analysis of camera feeds, and scanning of self-checkout behavior.
Startups to watch: Verkada, Everseen, and Solink.
Even more AI, now in GameDev
The video game market is growing and is already worth more than $190 billion: five times the size of the world's entire film industry. But there's a catch: video game development can cost hundreds of millions of dollars and often takes years. AI has the potential to dramatically speed up development and reduce costs.
For example, AI can reduce production costs by 94% for a single in-game character! In addition, according to one survey, 64% of developers are already using or would like to use AI in their work.
Analysts note strong year-on-year growth in headcount in most synthetic gaming companies. Investors should definitely pay attention to the area.
Startups to watch: Inworld AI, Modl.ai, Hidden Door, Anything World, Circle Labs, and Promethean AI.
Tired of AI? Too soon. Meet Humanoid Robots
The development of humanoid robots dates back to the 1960s and continues to this day. However, their use in the field (and in general any application beyond the laboratory) has not yet begun. As far as venture capital is concerned, the sector remains extremely niche, with only a few deals. Investment volumes are mediocre.
However, for the first time in many years, analysts have noted an increase in activity in 2023, which could be a sign of an imminent change in the environment. We recommend these companies to die-hard adventurers and admirers of Isaac Asimov's work.
Startups to watch: Agility Robotics, Fourier Intelligence, CloudMinds, Figure, and XPENG Robotics.
CVC Exodus
Corporate venture capital investment is declining globally, with 2023 results at the lowest level in five years. Low exit returns and cost optimization are among the reasons cited.
Corporations have dramatically changed their approach to innovation investment, focusing on strategic technologies. In other words, companies are now funding only what they really need, and cutting back on experiments.
Startups with products for the corporate sector may find it difficult to find customers.
One interesting detail. The most active CVCs are currently taking place in Japan.
Death of the Unicorns
There is a shortage of spare capital in the venture capital world today, which was not the case a few years ago. On the contrary, the number of startups is increasing. This imbalance creates unfavorable conditions for young (and not so young) companies, whose mortality rate in 2023 was one of the highest in recent times.
Unicorns are similarly at risk. According to the authors, many are on the verge of collapse and will face a drop in valuation, M&A, or outright closure shortly.
Curious about becoming an angel investor? You should consider joining our club. In the meantime, follow us on LinkedIn, Instagram, Twitter, and Facebook.