In the News: AI for Drug Development, Ensuring Chess Fairness, and Automating Google Sheets!
This week’s newsletter will recap Sanofi’s massive AI drug-discovery research deal and explore how Chess.com’s Strength Score ensures integrity in competitive chess. We will also introduce you to a startup that can bring live data to your Google Sheets!
Another One — Sanofi Signs Billion Dollar AI Drug-Discovery Deal with Insilico Medicine
Sanofi, a pharma giant from France, recently announced a research agreement for up to $1.2 billion with Insilico Medicine, an AI-powered biotech firm based in Hong Kong. The focus for the initial stages of the deal (a $21.5 million investment) will center around using Insilico’s Pharma.AI platform to discover up to 6 viable drug candidates for further development. In a “prove it” kind of contract, if Insilico and Sanofi meet their research goals, future investments totaling the billion dollar amount could spark an AI innovation hub for future drug development. This announcement is a landmark moment for AI’s growing maturity in the biotech space: AI systems have already shown breakthrough capabilities for predicting biological phenomena like protein folding (think DeepMind’s AlphaFold), which allows scientists to simulate drug interactions in computers rather than research labs. Think of the implications of such an achievement: if we can map the human genome and predict molecular interactions in computers, we can solve complex biological problems, finding cures and treatments for disease at the speed of computation. According to the article, in the US, it currently takes 10–15 years on average to develop a new medicine. Imagine how fast we can accelerate innovation in healthcare given viable AI systems! To drive home the excitement around this application of AI, I’ll close with one of my favorite quotes about AI and biology from Demis Hassabis, CEO of DeepMind: “Biology is emergent and hard to describe mathematically, but as math is a descriptive language for physics, artificial intelligence is the language for biology.”
Using Data to Promote Integrity — Chess Anti-Cheat Measures
Recently, the chess world went into uproar after current world champion Magnus Carlsen withdrew from a September 2022 tournament after a loss to rising star Hans Niemann. The community took this as an implicit accusation that Niemann had cheated (which Carlsen explicitly accused him of in a later statement), and the scale of the scandal prompted Chess.com to open an investigation into Niemann’s play. What I hope to focus on in this report is how crucial it is in data analysis to develop metrics that are relevant to your domain and not make sweeping conclusions based on statistical results alone.
Chess.com’s main metric in its quantitative evaluation of cheating is the Strength Score, “a measurement of the similarity between the moves made by the player, and the moves suggested as ‘strongest moves’ by the chess engine. In a sense, it is a measure of the accuracy of play.” The score is calculated by “actual statistical models and meant to be used [I interpret this as training the model] across multiple games.”
A key difference of this metric versus a player’s online ranking is that players can greatly over or underperform their ranking, but research has shown it unlikely that a player will significantly veer from their latent Strength Score for extended periods of time. By creating effective metrics like this that can be standardized, Chess.com is able to achieve its mission of providing an enjoyable, fair chess platform for players across all rating levels.
Finally, the report is clear about conducting further due diligence. Human analysts and players review the games to offer a qualitative lens, external factors such as browser behavior are considered, and a suspected player’s Strength Scores are benchmarked with other comparable players. However, data and ML models should serve as tools to drive better outcomes, not as a unilateral judge that can wrongly end a young upstart’s career!
Coefficient Syncs Google Sheets with Live Data
by Annika Lin
The startup Coefficient can integrate multiple data systems with Google Sheets. Users can create, share and automate live reports and dashboards, saving time spent on cross-referencing between multiple data sources. You can easily update your sheets with live data from sources like Google Analytics and MySQL, as well as customer relationship management (CRMs) systems like Salesforce and HubSpot. Can you imagine how your data analytics tasks will improve with this integral feature?
