You have a LOT of shortcuts!
@beardfm I thought you handled your “apology” video very professionally. Bravo! (As did the Raycast folk.)
I had not heard of Tuna, but I’ve been looking for something to fill that same style of Quicksilver.
I’ve been on Alfred since V1, and I’ll probably upgrade to a lifetime license next, but Tuna looks nice. I’m eager to try it out now.
Tuna looks very interesting! Trying now.
I had to pause the episode right at the end to do an emergency intervention ;-).
Right at the end @MacSparky was talking about Dot/OpenAI being hopefully more privacy preserving than Meta. I don’t think the evidence supports this. Recently in the world of research math, OpenAI appears to become aware of the work and direction that Tristan Buckmaster and Levent Alpoge were doing on Napier Stokes problems. OpenAI then used that knowledge to kickstart their own work in this area.
I don’t think this is an OpenAI problem. I think all of the labs have no respect for privacy, your data and anything related.
Perplexity is no better, the founder has said outloud that his users are a source of data to sell to advertisers.
I’m sure Anthropic also has it’s skeletons, I just haven’t noticed.
NONE of this players gives an ounce of respect for our privacy.
#Truth @mlevison. But at least we’re not hearing the horror stories we’re hearing about Muse. The way I’ve been navigating this is keeping personal stuff out of these systems, but I am using them for work. I look forward to a day when local models are good enough to throw aside these online models.
FWIW if you put the effort into it, some cloud providers will get give you real privacy with Open Weight models like Deepseek. An insanely ironic possibility. Meta has released open weight models that you could pay to run a service like Fireworks.ai. Then pick a harness /agent runner like Hermes. (I’ve not tested it at all.)
So I think a privacy focused version is possible now. The amount of effort to put together is daunting and I’m a recovering software developer.
I look forward to your experiments with your M6 mini and M5 laptop. My next machine is probably going to wait until the M8/9 generation. Our household normally runs on 5yr refresh cycles so my M3Max has another few years.