A cheaper token isn't a cheaper task.
A model with cheaper tokens can still cost more if it needs extra turns and more context to finish the job. Notes from a breakfast discussion with Earlybird in London.
A model with cheaper tokens can still cost more if it needs extra turns and more context to finish the job. Notes from a breakfast discussion with Earlybird in London.
We helped organise the {Tech: Europe} Agentic AI Hack in Stockholm. Almost half the submissions tried integrating condense, and one person got knighted.
4,500 builders applied to the Global Founder Sprint by a16z speedrun and fewer than 0.5% were selected. We're one of them. A week in San Francisco, 5 to 11 October.
On 2 September we hosted a private dinner in Stockholm for founders and senior leaders, with people from Lovable, Stripe, Einride, Epiminds and others.
We sponsored the London AI Cost Hack. The winning entry came second on cost and second on hidden-set quality.
We sponsored basedhack, an invite-only AI hackathon in Vilnius, and the proxy compacted 4B tokens of agent context over the weekend.