Astra on xhigh has a cost per task of $2.31 with an intelligence index of 53. Qwen3.8 Max has a cost per task of $5.41 with an intelligence index of 45. Pricing for GPT-6 Astra (xhigh) is $10.00 per 1M input tokens and $50.00 per 1M output tokens. Pricing for Qwen3.8 Max (0902) is $2.00 per 1M input tokens and $6.00 per 1M output tokens.
Obviously this is just one measure of all of this (and Qwen 3.8 Omni Flash isn't yet available), but I think this illustrates the point well. These relative task costs are pretty consistent across different analysts. Cost per token is arguably a useless measure at this point in most circumstances.
Je suis tellement heureux de constater que Carney bouge rapidement.
Au début des annonces des tarifs, plusieurs experts disaient que de se découpler des américains étaient impossible à court terme. Nos économies sont trop entrelacées pour changer de cap rapidement.
Je ne sais pas exactement ce que Carney a en tête , mais j’espère que ça va accélérer le découplage.
I'm in favor of Rank Choice Voting for this reason. Our system makes 3rd party votes hard or impossible, so instead of voting for who you like, you have to vote for who you dislike the least.
Let people vote for who they like the most without fear of their vote bolstering a candidate they really don't like. We should start exposing kids to RCV in schools. Now. It needs to not be "weird" or "hard" or propagandized as "not fair." Not fair is not being able to vote for your favorite person.
US biggest existential crisis is the fact that the opposition is always a monopoly.
So either you have a massive win, or a party will block the other forever out of spite, nothing can be accomplished and no mistake can be fixed. With more opposition groups the situation would change for good ("You don't want to make a deal with me, I will ask the third and fourth parties. The first team doing a reasonable move gets the prize").
As someone who does not own an F150 (or any truck), I find myself frequently wishing I had a truck, and either borrowing a friend with a truck, renting a truck, or doing some really unsafe things with my hatchback while muttering with my wife that our next car with be a truck.
If we had a truck, what % of the time would we be using it for things that can only be done with a truck? Very little, since it would also be our daily driver, but even if just 1% of the time, it's 100% worth having a truck over some sedan (or a hatchback!)
As someone who until last year has only had (mostly sports) cars or motorbikes, I now have a ute and it’s awesome to be able to do ‘ute stuff’ with it. Carrying toolboxes, transporting large/heavy/dirty cargo, collecting firewood, going off-roading and camping. I use it way more than I thought I would.
Likewise with my Jeep Gladiator. I traveled, hauled and camped with it more than I ever expected.
The bed is only 5 feet long but I'm thinking of taking out part of the back seat for even more room .
I also used 4x4 in 4H mode more than I thought on bad roads even though I didn't go truly off-road that much so no need for 4L. I did take advantage of the higher wheel clearance and the more rugged suspension and the 8-speed manual shifting is useful going down mountain roads.
Tbh as someone who does not drive or own a car regularly, I get by renting those U Haul vans (not trucks). It’s not expensive and it’s a hell of a lot more useful than a truck for doing the random heavy odd job.
Can't be more than 1% at this point. The cleanest cars on the road, and the only rough terrain they face is climbing kerbs to bypass queues in traffic.
It's wild to see all these people clowning on truck owners... and then say they are going to buy two 5090's for local AI even though they know it's less efficient than a cloud and don't necessarily need the privacy aspect of it... but want to do it just "cuz" . The hypocrisy here is amazing.
Even if OpenAI end up using 1 token for per task, if the token costs 1M$ , some people will find it expensive.
reply