The Free AI Era Is Over
What happens when AI stops feeling free?
We unpack the growing reality behind token pricing, disappearing “all-you-can-eat” AI plans, and why the economics of AI are starting to shift dramatically.
👉 dootrix.com/the-next-thing-n…
"I wonder if this is a bit of a taste of the future in that we're used to getting models as soon as they fit for public consumption. Maybe the slows as releases start getting more intelligent as these bigger models come out,"
"The UK, quite rightly, quite laudably, almost exciting saying, here's a stack load of money we're investing in the AI future. But at the same time, we're saying, as we all mature in this AI future, Flat is a new up."
"Coinbase tweet was very specific. And it talked about we can now do stuff in days that used to take weeks or months. We simply don't need the same number of people. AI is making more decisions or helping with the decision process, we also don't want to have managers anymore."
" Anthropic said a typical engagement starts with a small team working closely with the customer to understand what it is that their operations are required to do. Then their own developers will work with them to deploy Claude
Right. Okay. So it's a standard consultative play."
"do you have a policy about if should adopt a new version of Microsoft Word that they've released through Office 365? No, you automatically just take that 0.1 version number without blinking because cloud is understood and it's safe and it's gone through this a decade ago"
"the overriding Narrative around professional services for months has been it's dead or at least on life support. It's a matter of time But what's clear is that once you get beyond Being able to dabble is when you actually come to try and implement it You do need help"
"to be clear, we know the government doesn't actually have any money. That's why we give the government money. So they are spending 500 million pounds of our money"
"being left alone unattended. And that's the other thing that the Frontier models are really nailed. They're really good at doing long running tasks and these local models are not."