“It has been 2 years of the Gemini program and GDM. We now have come a great distance in that point with many efforts we should always really feel very pleased with. On the similar time competitors has accelerated immensely and the ultimate race to AGI is afoot. I believe now we have all of the components to win this race however we’re going to need to turbocharge our efforts.
Code issues most — AGI will occur with takeoff, when the Al improves itself. In all probability initially will probably be with a variety of human assist so a very powerful is our code efficiency. Moreover this must work on our personal 1p code. We now have to be probably the most environment friendly coder and Al scientists on this planet through the use of our personal Al.
Productiveness — In my expertise about 60 hours every week is the candy spot of productiveness. Some of us put in much more however can burn out or lose creativity. Numerous of us work lower than 60 hours and a small quantity put within the naked minimal to get by. This final group isn’t solely unproductive but additionally could be extremely demoralizing to everybody else.
Location — You will need to work within the workplace as a result of bodily being collectively is much more practical for communication than gve and so on. And, due to this fact you have to be bodily colocated with others engaged on the identical factor. We have to decrease reporting strains throughout nations, cities, and buildings. I like to recommend being within the workplace at the very least each week day.
Group — We have to have clear duty and group with excessive functioning teams with shared administration and know-how management.
Simplicity — Lets use easy options the place we are able to. Eg if prompting works, simply do this, don’t posttrain a separate mannequin. No pointless technical complexities (corresponding to lora). Ideally we are going to really have one recipe and one mannequin which might merely be prompted for various makes use of.
Excellence — whether or not it’s an eval or an information supply or a dashboard or a message in an inner Ul, please ensure that all of them work and all are good.
Velocity — we want our merchandise, fashions, inner instruments to be quick. Can’t wait 20 minutes to run a little bit of python on borg.
Iterate at small scale — we want a lot of concepts that we are able to check shortly. The easiest way to do that is small scale experiments till you possibly can ramp up and hopefully see rising benefit at scale. This is a wonderful validation. Working an excessive amount of at simply massive scale has a behavior of minor tweaking and overfitting to evals, checkpoint sniping, and so on. We want actual wins that scale.
No punting — we are able to’t maintain constructing nanny merchandise. Our merchandise are overrun with filters and punts of assorted varieties. We want succesful merchandise and [to] belief our customers.“