Kuona

Significantly more challenging than just 30s. Many many steps. 

Most steps not likely to see significant uplift from computational approaches as components are individually complex and very strongly interdependent in many ways - likely very challenging to gather appropriate training data to judge fitness of many combinations of variants and components, especially as intermediate outcomes are challenging to define for a reward function. 

Choosing simpler ribosomes erodes the knowledge accrued for other species like E. coli. 

The bar is reasonably low to a novel and publicised risk being included and described as a concern. There are concerted policy research efforts among well connected organisations that engage with most of these forums already. 

Trump wants to be unpredictable, but he may also be undecided. The military movements are substantial and costly if they are just for negotiation. They are also not necessarily needed for B2s. But Israel does not need them based on their fighter range. So perhaps they are for US base and defence support after all, but this would be primarily against missiles and drones. Attacking an Iranian manned aircraft would count, but there aren't many of these which would attack a US base, many are old or now destroyed. 

I noticed that those with more expertise in US politics predicted this substantially lower than me. I think I didn't consider how little time there is, and that there is not much progress - this wasn't considered in my model - which just said 'given that it hasn't passed yet' - many of these bills had more obvious signs of progress at this point. So I will weight this lower and halve again. 

Historically over 90% of bills passed at some point. But there is little time for this, a recess, and POTUS is busy elsewhere. VP indicated by August, reportedly, which is less ambitious. Grassroots concerns may complicate it. I think the base rate should be weighted very low as this administration is unusual and the time limit is what is important. 

The base expectation for how long a bill takes to pass  the second chamber having passed the first, as assessed by an LLM web search and collation of the data, is about 27 days (low confidence). Conditional on not having passed by now (day 28), there is roughly a 45% chance of passage before 1 July based on the distribution of the bills who had not passed at this time. 

So the naive prediction would be 0.45 x 0.9 = 0.405

But given the number of other factors reducing the likelihood, I will halve this. 

Update down due to narrowing lead in polls. Not updating based on others' interpretations of recent events (eg the ICE arrest etc) as low confidence in my ability to interpret these given unfamiliarity with US politics. 

Declining trend, in line with most of the reference class of previous presidents at this stage in their second term. 

Economic problems and foreign interventionism clash with wishes of some of his core base while being unpopular with many of his opponents too. 

Fairly stable trend during volatile time - I think only very extreme events could rapidly change things, and there's not much time for this, and some delay in polling.