if projections on exponential models. if the _ projections on exponential models. if the london school, if the problem is that— if the london school, if the problem is that even— if the london school, if the problem is that even a very small change in the inputs— is that even a very small change in the inputs reach to a very big difference in what you get out at the other— difference in what you get out at the other end, and there are some really— the other end, and there are some really critical things that we do not know _ really critical things that we do not know. we obviously do not know what level— not know. we obviously do not know what level it — not know. we obviously do not know what level it is going to peak, we do not _ what level it is going to peak, we do not yet — what level it is going to peak, we do not yet know what the effects in terms _ do not yet know what the effects in terms of _ do not yet know what the effects in terms of hospitalisations of two doses— terms of hospitalisations of two doses of— terms of hospitalisations of two doses of vaccine, prior infection and of— doses of vaccine, prior infection and of the — doses of vaccine, prior infection and of the booster dose are, and there _ and of the booster dose are, and there is— and of the booster dose are, and there is still debate about could this be — there is still debate about could this be a — there is still debate about could this be a slightly milder version? i do not _ this be a slightly milder version? i do not think it is to the point of triviality. — do not think it is to the point of triviality, but all of these are really — triviality, but all of these are really important. i am very careful, not _ really important. i am very careful, not all— really important. i am very careful, not all the — really important. i am very careful, not... all the way through the pandemic— not... all the way through the pandemic i resisted reaching for projection numbers except a short time period when it is more reliable into public— time period when it is more reliable into public domain because i do not think— into public domain because i do not think they— into public domain because i do not think they are reliable and you cannot— think they are reliable and you cannot put enough caveats on them. lets cannot put enough caveats on them. lets look—