AI is all the rage. BenevolentAI seeks to use this rapidly advancing technology to transform how medicines are discovered, developed, tested and commercialized. It has designed a technology platform to help drug developers to sift through mountains of medical data (including clinical trials and academic paper) as it ultimately makes recommendations. The platform viability hasn’t been proven on the market and they may be headed toward a “down round.”
Forbes Sam Shead reports on Baroness Joanna Shields, a British-American technology business veteran who serves as BenevolentAI group CEO. Shields is a mover and shaker in high circles in British politics and business having been made a digital economy advisor to David Cameron, not to mention a Life Peer in the House of Lords in 2014. She was even appointed the Prime Minister’s Special Representative on Internet Safety. Prior to ongoing mingling at the upper-echelons of Anglo-Saxon society, she ran with the top corporate chiefs in tech from RealNetworks to Google, AOL and Facebook not to mention some prominent start-ups. She is beautiful, charming, intelligent, charismatic and fortunate. But even that isn’t enough when the market turns unless she makes some bold, intelligent and shrewd bets.
Now Forbes is reporting that the hot drug discovery-focused AI venture is in seek of fresh money at a deeply discounted valuation reported The Times. This means the once high flying star is now at risk of losing its “Unicorn” status which requires at least a $1 billion valuation.
TrialSite News took a brief look at their business.
Formed in 2013 they have raised an enormous amount of cash—according to website CrunchBase a total of $202 million. The company must be burning through cash. They have a reported 184 self-reported employees on professional social network site LinkedIn. It should be noted most professionals maintain their LinkedIn profiles. Much of the reviewed titles will command considerable salaries in markets such as London and New York where they operate. Start-up tracker Owler reports 300 employees. We suspect they are right at about 200.
We suspect they are spending at least $30 million per annum if not up to $40 million annually or $3.3 million per month. The Financial Times reported that it had not yet to draw down the full 84.1 million UK pounds it raised last year and recent filings in the UK revealed they still had 20.5 million pounds to come in at the end of 2018.
So adding the figures above and the 19.9 million pounds it had on its balance sheet it would have available 40.4 million pounds readily available for the coming year. The Financial Times reports this would be enough for the operating loss of 32.6 million pounds. They had wiggle room at last year’s spending levels and TrialSite News believes our estimates are in the ballpark.
In an interview with the Financial Times, a BenevolentAI spokesperson downplayed certain news and pointed to their AstraZeneca deal as revenue sources not to mention “other such deals in progress.”
That recent news covered by the Financial Times didn’t bode well for their second largest investor, PatientCapitalTrust, in which BenevolentAI is its second-largest holding at around 85 million UK Pounds.
BenevolentAI will probably be sold at a deep discount. It is ahead of its time in some ways. Yes, there is a lot of AI infused into drug discovery now; yes it’s a great idea; yes we would personally serve the Baroness tea and biscuit just to hand out with her—but the unfortunate problem here is that the BenevolentAI’s model must come in alignment with where the market is actually at. Let’s not forget the IBM Watson story. They spent hundreds of millions; if not more—to position and market AI services. They paid dearly for this. In the long run—or even in the intermediate run—there is no stopping AI—that train has left the station. But markets take their time to adopt technologies in a way that they can align their core business challenges with optimal ways to consume the new technologies.
Our premise is that BenevolentAI cannot command huge premiums from drug developers because there is, frankly, a lot of competition out there. See infographic in the link. There are many ways to consume AI services for augmenting drug discovery. Even if BenevolentAI’s technology is superior, and we certainly don’t doubt Baroness Shield’s judgment—theirs is probably better than many others. But that doesn’t mean the markets are ready to consume these algorithms at the velocity, pace and scale (and associated economics) that the investors desire.
A Conservative Bunch
Drug discovery purchasers are generally conservative, risk-averse and will seek a return on investment (ROI) and performance-based milestone payments. The bigger checks will require significant results.
More than likely they (BenevolentAI) will need to adjust and align themselves with the actual market demand for their products and services. Their model isn’t Amazon where they must run at deficits for a decade-plus because at the end they will dominate the world.
We don’t see the drug discovery space involving AI as a winner-take-all market. The markets don’t operate that way and different pharma are very careful in who they align with when it comes to matters of their core IP formation.
Or Think Roll-Up?
Downsizing, laser-like focus on high-value add contribution and relentless execution will probably be their future within 24 months or less. On the other hand, the Baroness could take some of that capital and start consolidating and rolling up the many AI plays focusing on drug discovery—that could be interesting.