1. 13

  2. 5

    I’m not an expert on the subject (the first part of the post has lots of jargon) but the second part that begins at the chapter “What just happened?” is very interesting and understandable. Thanks for sharing.

    1. 2

      Yep, that’s what was especially interesting to me too. I found a lot of food for thought there, with implications potentially much wider ranging than purely for the biotech field — and at the same time a glimpse behind the curtain on some current state of affairs in the biotech science community.

    2. 2

      If you think I’m being overly dramatic, consider this counterfactual scenario. Take a problem proximal to tech companies’ bottom line, e.g. image recognition or speech, and imagine that no tech company was investing research money into the problem. (IBM alone has been working on speech for decades.) Then imagine that a pharmaceutical company suddenly enters ImageNet and blows the competition out of the water, leaving the academics scratching their heads at what just happened and the tech companies almost unaware it even happened. Does this seem like a realistic scenario? Of course not. It would be absurd. That’s because tech companies have broad research agendas spanning the basic to the applied, while pharmas maintain anemic research groups on their seemingly ever continuing mission to downsize internal research labs while building up sales armies numbering in the tens of thousands of employees.

      The next time someone is defending the necessity of pharma patents and absurdly high drug prices, you can be certain they are shilling for entities that are trying to kill you for a profit, and act accordingly.

      1. 1

        As someone who saw Mohammed give a talk about this several months earlier and was also one of the people who asked this to him basically point-blank at NeurIPS last week, I also am somewhat surprised. I think the important takeaway is this is in some sense incremental work building on what many people in the space are doing without Alphabet’s resources. I think the pessimistic angle is that if the main barrier to headline-making results is computational resources Alphabet can scoop you. The optimistic angle is that very few problems are strictly compute-limited.

        1. 2

          For me personally, I feel the implications sketched in the second part of the article may be much wider ranging than purely for the biotech industry. I see it as a very interesting case study of ML (and various approaches to it) vs. the world/science generally, with the biotech field being just a somewhat accidental specimen on which the dynamics can be observed. That kind of generalization and a glimpse of a wider perspective is what makes the article especially powerful for me.