It’s like an upside down freemium model - try out our basic self driving product, which is (now) the worst in the market, so you’ll convert to the premium FSD offering.
1,646 karma · joined February 14, 2022
Get in touch at hello@tombedor.dev, or at https://x.com/middleagedc0der
It’s like an upside down freemium model - try out our basic self driving product, which is (now) the worst in the market, so you’ll convert to the premium FSD offering.
This would be my critique of MCP-specific security implementations. I think robust tools for this already exist, and in general AI API calls can and should be treated like any other RPC call.
Looking at the post again, I think I agree that calling Claude Skills overengineered is too harsh. I think Skills is definitely an improvement over MCP.
However I still think it's a generally a mistake to put useful commands and documentation in AI-specific files. In my opinion a better approach is to optimize the organization of docs and commands for human usability, and teach the AI how to leverage those.
I do use Claude Skills, but only to wire up just commands. I wrote a little package to do this automatically: https://github.com/tombedor/just-claude
Thanks for flagging! I just pushed an update that should fix this
I wrote a bit on the topic here: https://tombedor.dev/make-it-easy-for-humans/
Consumer hardware chips will be plenty powerful to run “good enough” models.
If I’m an application dev, do I want to develop something on top of OpenAI, or Apple’s on device model that I can use as much as a I want for free? On device is the future
Nothing against people who like them, to each his own. But the throughput of quality programming out of HBO has dropped off a cliff through it's multiple changes in ownership.
Unless there’s an accusation that Eric / Core is violating license, I don’t see a lot of merit to Rebble’s position.
Is this the stat you have issue with? Or is the contention that a pet on a city street at 11:40 PM is not highly at risk of being run over by a human driver?
If "we'll have too many cars on the freeway following the speed limit" ranks as a serious concern, I think we've really lost the plot.
I recently drove by a fatal accident that had just happened on the freeway. A man on the street had been ripped in half, and his body was lying on the road. I can't imagine the scene is all that unlike the 40 thousand other US road deaths that happen every year.
As a driver I'm willing to accept some minor inconvenience to improve the situation. As a rider I trust Waymo's more than human drivers.
The political climate is VERY suspicious of autonomous vehicles, but they most serious incident I can really recall was the recent one where a car ran over a cat. You can see the reaction here: https://www.reddit.com/r/cats/comments/1omortk/the_shrine_to...
If the biggest black mark against the company is running over a cat on the street at 11:40 PM (according to Waymo, after it darted under the car), I feel pretty good.
Things like, “get xyz a birthday gift”, and it finds the contact and sets the reminder for a few weeks before the birthday.
The source data is calendar and contact info, but the nice thing about local models is that no data gets sent to providers, and the app can be cheap
So basically, I'd love to have my storage provider give me a vector search API, which I guess is what Amazon S3 vectors is supposed to be (https://aws.amazon.com/s3/features/vectors/)?
Curious to hear what experience people have had with this.
aka the entire point of the exercise. The innocuous components are there so that the Dept of Defense can claim that it's those minor items the press is objecting to, without having to defend the actual substantive policy change.
The pretext is really falling away.
Terrible precedent aside, how could Disney think that capitulating here will result in anything other than more attempts to control their programming in the short term?