The excitement surrounding artificial intelligence (AI)’s potential to transform healthcare has led to substantial investments and a proliferation of startups eager to harness the power of machine learning and data analytics. Indeed, the potential of AI in biotech is a true game changer and it is in the background of several trends we thought you should keep an eye on in 2024. From drug discovery to drug repurposing, there’s no doubt it is an essential tool for the industry. However AI in biotech is experiencing a dowturn, and it could seem the bubble is bursting.
As the industry matures, the initial wave of optimism is being tempered by the realities of complex biological systems and the challenges of translating AI-driven discoveries into clinical successes. Recent events, such as the $688 million merger between Recursion Pharmaceuticals and Exscientia, two of the most important players in the field, have brought the industry’s challenges into sharp focus.
In this context, a question arises: Is the biotech AI sector in the midst of a market bubble that is beginning to burst, or are we witnessing a necessary correction as the industry moves from hype to a more sustainable, mature phase?
Table of contents
Setting the stage: The Recursion-Exscientia merger and BenevolentAI’s decline
The merger between Recursion Pharmaceuticals and Exscientia is one of the most significant moves in the AI-driven drug discovery industry. This deal will see the U.S.-based Recursion absorbing its U.K. counterpart, Exscientia, creating a “full-stack technology-enabled small molecule discovery platform” powered by AI. This new entity will house a portfolio of 10 clinical programs already in testing.
The strategic goals behind this merger are clear: to combine Recursion’s extensive capabilities in biological exploration and translational research with Exscientia’s precision in chemistry and newly commissioned automated small molecule synthesis platform. The expected outcome is a more efficient drug discovery pipeline that could potentially revolutionize how new therapies are developed, leading to broader availability of high-quality medicines at lower prices for consumers.
Ram Srinivasan, managing director at JLL Consulting, sees this as complementary technologies and expertise coming together to create value. “This partnership merges advanced biological research with cutting-edge chemistry, potentially speeding up and improving the drug discovery process.” According to him, this move aligns with broader trends in the industry where companies are consolidating to enhance their technological capabilities and streamline operations.
Meri Beckwith, co-founder of Lindus Health agrees. “It reads to me mostly as a strategic consolidation. There are interesting synergies with Exscientia providing target identification and Recursion providing drug design, and the usual operational efficiencies too. Additionally, the combined entity has more shots on goal across their pipeline. I think the merger is therefore great news for both companies and I’m excited about the potential!”
While this merger is certainly a strategic move when considered on its own, it does contribute to the skepticism that still casts a shadow over the impact of AI in drug discovery. This merger comes in a context of uncertainty regarding AI as the two companies’ market value experienced a sharp decrease.
Indeed, Yuguang Mu, founder of Magmole, a Singapore-based company specializing in AI-accelerated drug discovery, thinks the context of the merger shouldn’t be overlooked. “The merger of Recursion and Exscientia may have some complementarities but both companies still face many pressures. Also, it remains to be seen whether the proposed ‘end-to-end drug discovery platform’ will be the outcome since both teams need to spend a lot of time sharing knowledge, processes, and best practices in receptor, protein, and RNA targets.”
According to Artem Trotsyuk, partner at Longe VC, it’s a bit of both. “The merger between Recursion Pharmaceuticals and Exscientia appears to be both a strategic consolidation and a response to market pressures. It combines complementary AI approaches, potentially accelerating drug discovery. However, it also reflects the need for scale and efficiency in a challenging market.”
Recursion and Exscientia are not the only ones to have faced setbacks, as BenevolentAI, another major player in the field, has seen its market value plummet. Considered one of the most promising players in the field, BenevolentAI has faced a series of difficulties, the most notable being the failure of its atopic dermatitis candidate, BEN-2293, in a mid-phase clinical trial. This drug, which was developed using the company’s AI platform, failed to outperform a placebo in improving eczema symptoms, raising questions about the effectiveness of the company’s AI-driven approach.
In this context, investors and industry observers are beginning to question whether the AI promises of faster, cheaper, and more efficient drug discovery can consistently deliver tangible results. In other words, is the AI biotech bubble bursting?
Decoding the turbulence: AI bubble bursting or market correction?
While the term “bubble” implies a dramatic and potentially destructive collapse, a market correction describes a necessary phase of realignment as the biotech AI sector adjusts its expectations to align with real-world results.
Is the AI bubble bursting?
A recent episode of The New York Times’ Hard Fork podcast offers some answers to this question by comparing the current AI crisis to the dot-com bubble.
The dot-com bubble was a period of extreme growth in the value of internet-based companies, primarily in the late 1990s. Fueled by speculative investment in companies associated with the burgeoning internet, the bubble saw stock prices skyrocket, often without regard to the companies’ actual profitability or business models.
By March 2000, the Nasdaq reached its peak, driven by irrational speculative investment. Companies with no clear revenue streams or sustainable business models were achieving astronomical valuations purely based on their association with the internet. The bubble burst in 2000 when investors began to realize that many dot-com companies would never turn a profit. This led to a massive sell-off.
Similarly, in the AI sector, there is a prevailing expectation that AI will inevitably lead to breakthroughs in fields like healthcare, often leading to inflated valuations. Just as the dot-com bubble was fueled by speculative investment in internet-based companies, the AI market is currently experiencing a similar influx of capital.
However, the comparison has its limits as Hard Fork’s podcast host Kevin Roose noted that the companies that are going under now don’t have the same status as the ones that did during the dotcom bubble. “Most of the companies that are sort of washing out, these mid-level AI startups, are private companies. They’re not being traded on the stock exchange. The companies that are public, the Metas and Amazons and Googles of the world, have these massive piles of cash that they’re sitting on. And they have these other businesses that are kind of subsidizing their investments in AI.”
“So, will NVIDIA be able to sustain its current stock price? I have no idea. But I think the idea that these companies are all going to come crashing down because they spent too much developing large language models is a fantasy,” said Roose in the podcast.
If the AI biotech bubble were to burst, we might see a significant number of AI-focused companies either go bankrupt or be absorbed by larger, more established firms. This would particularly affect startups that have been heavily reliant on speculative funding without a clear path to profitability.
However, just as the internet did not disappear after the dotcom crash, AI is likely to continue evolving and becoming more integral to various industries. The dot-com crash, despite its immediate negative impact, ultimately led to the maturation of the internet sector, with surviving companies like Amazon and Google going on to become some of the most valuable companies in the world. Similarly, an AI washout could lead to a more sustainable and focused industry, where only companies with viable, proven technologies survive.
In the Hard Fork podcast, the concept of “sea change” was used to explain the AI situation. This concept describes a fundamental shift in how things operate within a particular industry. Volatile market valuations, company mergers, and failures should not be mistaken for a mere bubble burst that might lead to the demise of AI’s role in various sectors, including biotech.
Instead, this turbulence could represent a deeper, more fundamental transformation in how AI is integrated into industries and how these industries evolve around this technology. Despite the current setbacks and potential failures among AI companies, the underlying AI technology will continue to advance and transform industries over time, including biotech.
AI experiencing a necessary market correction
The biotech AI sector has been the subject of enormous expectations, with promises of revolutionizing drug discovery by drastically reducing time and cost. However, as Edward Tian, chief executive officer (CEO) of GPTZero, points out, the industry is currently grappling with a significant gap between these high expectations and the realities of what AI can deliver.
“I think this whole situation signals a market correction more so than the AI biotech bubble bursting. Largely, what we are dealing with is unmet expectations in the AI biotech sector. Big promises have been made about how transformative this technology will be, and expectations have been high. But, though there have certainly been significant advancements, those expectations haven’t quite been met, and that is influencing the market,” he said.
Nicholas Rioux, chief technology officer (CTO) of Labviva, echoes this sentiment. “There is a tremendous amount of interest and investment in AI biotech applications, but in a hype period, companies tend to lose focus, and the result is consolidation. There are enormous applications for AI in this sector, but in order to reap the value and justify their valuations, biotech and life sciences companies must focus on real applications of AI that either remove tasks from humans or empower humans to be more impactful. On the flip side, AI applications promoting vague benefits often cannot be quantified and will lead to missed valuation goals.”
Trotsyuk attributes the current challenges in the AI space to several factors. “First, AI-driven discoveries take longer to progress to clinical stages than initially expected. Second, high cash burn rates for AI research and development and broader economic pressures like rising interest rates are straining resources. Finally, there’s a misalignment between market expectations for short-term returns and the inherently long-term nature of drug discovery.”
Mu agrees with Trotsyuk that this downturn is due to a mismatch between a lack of big scientific breakthroughs through AI in drug discovery and overly high expectations from AI. “It could be a broader market correction. AI is a tool for complex data analysis but may not bring breakthroughs in sciences, especially in medical sciences. People tend to expect too much from AI in drug discovery and it turns out AI is not omnipotent. As expectations and breakthroughs seek to find a new lower equilibrium, lower valuations are likely.”
The initial overenthusiasm for AI’s potential is now likely being tempered by the practical challenges of applying these technologies to complex biological systems. This is a common trajectory in emerging technologies, where a period of inflated expectations is followed by a correction phase, leading to more realistic and sustainable growth.
And what if these downturns were neither linked to the AI biotech bubble bursting or a market correction? According to Beckwith, these valuations are largely driven by specific events at each company.
“The argument that there is no premium on an asset’s value just because it was ‘designed’ by AI cuts both ways. Any decline in share price will probably be driven by the performance of the lead assets and doesn’t necessarily reflect investor confidence in the process or models that designed them. Companies like BenevolentAI suffered from poor readouts from their lead candidates and an unclear commercial strategy. Meanwhile, other AI companies like Tempus have fared much better, doubling its share price since its initial public offering,” explained Beckwith.
The road ahead of AI in biotech amidst uncertainty
Despite these challenges, there remains a strong belief in the long-term potential of AI in biotech. Henry Levy, president of Life Sciences and Healthcare at Clarivate is optimistic about AI in healthcare, although unanswered questions remain. “Companies like Exscientia, with AI-designed drugs in trials, and innovations like Google DeepMind’s AlphaFold 3, predicting protein interactions, show how AI can accelerate research and discovery. But a key question remains: Are regulators ready for these rapid advancements? Fully integrating AI into healthcare promises significant strides toward a healthier future for all.”
Srinivasan might be even more optimistic. “We are at an inflection point where AI in biotech is transitioning from potential to practical application. This is the perfect moment for visionary companies to lead the charge. The gap between promises and current capabilities is closing rapidly. In my experience, this is when true breakthroughs happen.”
“We are on the cusp of a new era in drug discovery, and these fluctuations are growing pains on our path to reshaping global health. The challenges we face today are laying the groundwork for tomorrow.”
According to Beckwith, there is no reason to worry about these setbacks and AI is on track to deliver its promises. “The promise of AI was never just that about AI-designed drugs having a higher likelihood of progress to the clinic. It’s also about being able to create broader, more diverse pipelines of assets and bringing them to [the] Investigational New Drug (IND) [stage] more cost-effectively. AI is delivering on that promise.”
Tian and Rioux agree that AI will not disappear from the biotech sector and expect it will become more prominent as larger advancements are made in the future. “I see the value of AI for many life sciences and biotech applications, particularly where AI applications start with hard cost savings and other metrics that demonstrate efficiencies of scale so investors and customers can understand the value of their investment,” said Rioux.
Beckwith identifies that recent progress has shown that AI can effectively support target identification and small-molecule drug design but there is still a lot of room for improvement. “Despite this progress, we’re a long way from a ‘hands-off’ approach, where scientists can sit back and LLMs (large language models) will do the work for them. There is an opportunity for AI to further optimize chemistry in drug discovery, as the existing AI tools in this field are still in their early stages of development.”
Most importantly, Beckwith thinks there is an opportunity for AI that hasn’t been explored as much as drug discovery. “Most of the investment and deployment of AI so far has been in drug discovery or preclinical stages. However, the great majority of drug development costs are incurred in later-phase clinical trials in humans, where AI adoption is very low so far. This suggests a lot more potential for AI to have an impact. Pharma is, however, a very late adopter of new technology, particularly when it comes to clinical trials, so I’d expect to see slow change over the next decade.”
Although it might look like the AI bubble is bursting, the industry experts we reached out to are considerably more optimistic. The journey of AI in the biotech industry is still long and there might be a few bumps on the road but let’s keep our hopes up that it will help bring new solutions to market more efficiently.
But it would be wiser to temper our expectations, according to Trotsyuk. “We’re in a ‘trough of disillusionment’ phase. While the technology holds significant promise, it’s not yet fully mature. In the short term, we should expect incremental progress rather than revolutionary breakthroughs. The long-term potential remains substantial, but realizing it may take longer than initially thought.”
Partnering 2030: The Biotech Perspective 2023
