I like how he explain API basics to the developers🧡🧡
@MaciejStachowiak8 ай бұрын
we've meet at NVIDIA GTC I guess - very cool video here :)
@rishabincloud9 ай бұрын
Craig! 🙌
@brian_ball5 ай бұрын
The GOAT of devrel!
@kasper3699 ай бұрын
amazing stuff,
@jitxhere9 ай бұрын
Love cloudflare at 3:50 the slide says images and audio... Do you have policy against storing images at R2??
@danvilela9 ай бұрын
Man this was good! 😆👏
@ritam_chakraborty9 ай бұрын
Haven't I already seen you explaining twilio staff? 🤔
@Innesb8 ай бұрын
Yes, I recognised Craig too.
@masternmargarita4 ай бұрын
Keeping it real
@adriablancafort7 ай бұрын
I love cloudflare
@craigsdennisАй бұрын
Same!
@InSaneRoGer1120039 ай бұрын
awesome tech
@GabbyShires8 ай бұрын
Go Craig!
@ElenaBG38 ай бұрын
Never knew R2 stands for Robust and Reliable…🎉
@DavidD.j.GaffordАй бұрын
I was actually thinking about sum how sum way to use A.I.to wipe out high data usage on cell phones,I live off a cell phone have 3 of them and there my life my money and my way of productivity 2 wipe out data usage an turn 5G into 500 G.
@shafu_xyz9 ай бұрын
cooool
@CookerSingh9 ай бұрын
Workers A.I biggest downside is that the models can't remember the previous prompts asked by the user. It doesn't maintain conversational memory i guess becoz, its serverless.
@bringbackwindowsphone9 ай бұрын
That's the exact same as any other AI API/SDK though, including Anthropic and OpenAI. I was surprised to find you need to pass the conversation every request.
@akshatsethi63849 ай бұрын
This is a L comment
@CookerSingh9 ай бұрын
@@akshatsethi6384 don't be a double L then.
@CookerSingh9 ай бұрын
@@bringbackwindowsphone So it's on the model part whether i use token based approach for context retention (as in openai) or just dumped the whole conversation again and again. Why they can't just do that in the workers api for all the models, also I think if they give out function calling support as a proxy in their workers api for all models will definitely make this a great service.