David Kim

If I had to use an analogy, the gap is closer to that between Mercedes-Benz and Tesla than between two companies competing on the same level.

It's true $ABCL will eventually run into limitations This is not to say that their platform is poor or that the business is destined to fail. However, I often see people drawing comparisons between this company and Palantir, and I do not believe the comparison is appropriate. The level of impact, strategic importance, and long-term potential that Palantir possesses is simply not something this company has demonstrated. Their platform may be valuable within its specific domain, but it does not currently exhibit the same breadth of influence or transformative potential.

Platform validation absolutely matters. The real question is what kind of company investors are underwriting $ABCL to become: a $GMAB-like mid-sized biotech, or a $LLY-like large pharmaceutical company. If the expectation is the former, then ABCL’s current capabilities may be enough to justify the investment case. But if the expectation is the latter, then the company still has significant strategic gaps, especially in the areas I mentioned earlier.s, especially in the areas I mentioned earlier.

While most of your points are factual, the problem is the way you’re inaccurately framing it to fit a bearish narrative. This is a strawman. The bull case was never that ABCL discovered SARS-CoV-2 or independently commercialized bamlanivimab. The bull case is that ABCL demonstrated world-class antibody discovery speed and generated a molecule that Lilly deemed worthy of global development and deployment which is exactly the type of platform validation investors like me care about. Nobody is investing in ABCL today because they think ABCL created the spike protein. I appreciate the bearish response as it’s always important to look at an investment from both ends, and while your statements are factual, context and framing matter

$ABCL This narrative looks like a classic stock pumper story that stretches AbCellera’s real achievement far beyond what it actually proved. It is true that AbCellera rapidly identified antibody candidates during the Covid pandemic. However, using that episode to present AbCellera as a leading antibody therapeutics company is a serious exaggeration. 1) bamlanivimab was a monoclonal antibody, and the target itself was not newly discovered by AbCellera. At the time, the relevant target was already clear: the SARS CoV 2 spike protein and its receptor binding domain. AbCellera’s role was not to discover a new disease target, but to rapidly screen human antibody candidates from the blood sample of a recovered patient and identify antibodies that bound to that known antigen. 2) the core work of clinical development, manufacturing, regulatory execution, and commercialization was led by Eli Lilly. The case depended on the scientific foundation and antigen design from NIAID and VRC, Lilly’s development and manufacturing capabilities, and the unusual regulatory environment created by the pandemic. Therefore, it is inaccurate to frame this as proof of AbCellera’s independent therapeutic development capability. Moreover, the emergency authorization for bamlanivimab as a standalone treatment was later revoked because of concerns related to viral variants. As a result, this case is better understood as evidence that AbCellera can rapidly discover antibody candidates, not as proof that its platform has demonstrated sustainable superiority in drug development or that the company has independent clinical and commercial leadership. In short, AbCellera’s Covid achievement was real, but the pumper narrative around it is misleading. It showed speed in antibody discovery, not dominance as a full antibody therapeutics company.

If $ABCL truly wanted to become a broad drug discovery platform, it should have invested deeply in the upstream layers of biology and target discovery, rather than remaining focused on rapid antibody discovery and generation. But at this point, the company still appears concentrated on being a fast and efficient antibody discovery engine. That raises a very simple question: if management is truly ambitious, why has AbCellera not invested more aggressively in these areas despite holding such a large cash balance? The longer this continues, the wider the gap will likely become between AbCellera and the existing players that are already building deeper capabilities across disease biology, target discovery, translational research, and internal pipeline development.

@Da_Davidut do you think a management team who have gotten this far doesn’t have the ambitions to go higher?

AI-driven target selection is not what $ABCL has actually demonstrated.🤨 What AbCellera has demonstrated is AI/ML-assisted antibody discovery and optimization against targets that are already defined, often by its partners. That is a valuable capability, but it is a very different layer from understanding disease biology, selecting the right therapeutic target, and proving that target clinically. Calling $ABCL an “AI biotech” overstates the nature of the company. Its core business is antibody discovery and development, with machine learning used as an enabling tool rather than as the central engine for solving disease biology.

$GENB In AI-driven drug development, the company that is clearly the most advanced from a clinical standpoint is Generate Biomedicines. They are not simply a biotech company that screens existing drug candidates. Rather, they are a generative biology company seeking to design protein therapeutics by combining artificial intelligence with biotechnology. Importantly, they have already advanced GB-0895, a long-acting anti-TSLP antibody for severe asthma, into global Phase 3 trials. GB-0895 is particularly notable because it aims to enable dosing once every six months compared with existing therapies. If its efficacy is proven, this could give it a clear advantage in terms of patient convenience and market penetration. On top of that, Generate Biomedicines has collaborations with major pharmaceutical companies such as Novartis and Amgen, while NVIDIA’s venture arm, NVentures, has also invested in the company and holds an equity stake. $RXRX $ABSI $ABCL $OABI

From my research, $SATL stands out as a highly compelling company relative to other players in the market. The key point is that it does not necessarily need to be judged only through a direct comparison with $PL. This is a market where full monopolization by a single player is unlikely. What matters more is whether a company can carve out a valuable niche, operate efficiently, and generate attractive margins within that segment. From that perspective, $SATL looks like a very strong business.

If $ABCL is truly an AI biotech company, where is the evidence of a serious AI infrastructure layer? Where are the strategic partnerships with NVIDIA, AMD, Oracle Cloud, Google Cloud, AWS, or other large-scale compute providers? What AbCellera has shown is not an AI-native drug discovery model. It has shown an antibody discovery platform that uses machine learning as a supporting tool to analyze experimental data, rank candidates, and improve screening efficiency. That is useful, but it is not the same as building foundation models for biology, scaling generative drug design, or using AI as the central engine for target discovery and disease biology. In other words, $ABCL may be a strong antibody discovery company, but calling it an AI biotech overstates the nature of the business.

Isomorphic is already moving in a direction that attempts to go beyond $ABCL’s structural limitations. AbCellera’s core strength lies in discovering and optimizing antibody candidates against targets that have already been defined. This is clearly a meaningful capability. However, the highest value in drug development is not determined by the ability to find a “good antibody candidate” alone. The more fundamental value lies in identifying which target should be pursued in a specific disease, converting that target into a viable therapeutic candidate, and ultimately proving efficacy in the clinic. In other words, AbCellera has demonstrated strength at the candidate discovery and optimization layer, but it has not yet sufficiently proven its ability to understand disease biology and independently discover and validate effective targets. Isomorphic, by contrast, is attempting a much broader approach that connects disease biology, protein structure, binding pockets, and molecular design. This distinction matters. If AbCellera is focused on answering the question, “Which antibody should be used to attack a defined target?” then Isomorphic is closer to asking, “What should be targeted in the first place, and what kind of molecule should be designed to attack it?” Ultimately, companies that rise to the top tier of biotech cannot rely solely on candidate-generation capabilities. They must understand disease biology, identify valid therapeutic targets, translate those targets into drug candidates, and ultimately prove clinical efficacy. In that sense, AbCellera may be an excellent antibody discovery platform, but it has clear limitations as a broad biological problem-solving platform or as a next-generation major pharmaceutical company candidate. Based on its current structure, AbCellera may still occupy a meaningful position within biotech, but it cannot stand shoulder to shoulder with the companies at the very top. That is the reality.

$ABCL limitations are quite clear. They are not so much a company that independently discovers and validates core disease targets, but rather one that identifies and optimizes antibody therapeutic candidates against targets that have already been defined or selected by its partners. To be clear, this is still an important capability. However, in drug development, the greater value does not come simply from “finding a good antibody.” It comes from first understanding which target should be pursued in a given disease, then creating a therapeutic candidate against that target and clinically proving that it works. Ultimately, the companies that can grow into major pharmaceutical players are those that can do both. First, they must be able to understand disease biology and identify valid therapeutic targets. Second, they must be able to translate those targets into actual drug candidates through discovery and optimization. AbCellera has shown strength in the latter — antibody candidate discovery and optimization — but it has not yet sufficiently proven the former: target discovery and disease program design. Therefore, it is more accurate to view AbCellera not as a broad biological problem-solving platform, but as a specialized platform for generating drug candidates within the specific modality of antibodies.

@tyler_bosserman Yes, that differentiation matters. My concern is that many posts are heavily overstating AbCellera’s Covid-era achievement, and that can mislead people about what the company actually proved. That was the reason I wrote the post.

The work they are doing in GPCRs is what is really difficult and what should help them stand out from the crowd. Recently Carl Hansen was asked if abcl575 which targets ox40l highlights their capability and he basically said that there isn’t anything particularly hard about it. I agree - it isn’t something that differentiates them. But ion channels and GPCRs along with potentially ADCs and TCEs are what should will make them stand out, assuming it works. Abcl635 is a small proof point. We’ll know more in Sep.