Screen shot from Apple’s 1987 video about Knowledge Navigator. The resolution of the old video is poor, so I asked ChatGPT to redo the image, then I scaled the image up using Adobe’s Firefly tool. Click here for high-resolution version.
If there was a moment when I became a lifelong Apple groupie, it was 1987, when I attended an Apple marketing event in Charlotte, North Carolina. The event was to promote Apple products for publishing. Apple, along with Adobe, already knew in 1987 that their products would become essential in the publishing industry. During my career in publishing, I spent many hundreds of thousands of dollars on Apple and Adobe products for the newspapers I worked for. The costs were easy to justify, because the new tools for digital publishing were so much less expensive than the systems they replaced.
At that event, Apple showed the Knowledge Navigator video. It was like a dream — Steve Jobs’ dream. Could it ever happen? If it could happen, how long would it take?
Ironically, Apple fell behind on AI systems. Siri was an industry joke. But all of a sudden, everything is now coming together to make Steve Jobs’ dream a reality — an AI assistant with a video avatar, with access to your telephone and email, knowledge about who you are and what you do, the ability to do real-world research, and even the ability to generate pictures and video.
The rumors are that Apple will release a new version of its HomePod this year, with a screen. It is thought to be closely integrated with a new Siri powered by OpenAI’s ChatGPT. I am hoping that Apple will surprise us by including a talking video avatar for Siri, with your choice of avatars (whom you might or might not name Siri). A talking video avatar for AI is going to happen soon. It would be ever so cool if Apple did it first. And if Apple does it in 2026, then it will have taken 39 years for Steve Jobs’ dream to become a reality.
ChatGPT’s imagining of what an Apple video avatar might look like.
Note on downtime
There was some sort of problem yesterday (January 15) with this blog’s hosting service (GoDaddy). The blog was either down or partially down from around 2 p.m. until 9 p.m. EST. GoDaddy was very vague about what caused it. But it may have had something to do with the the fact that Monday’s post “One of ours, all of yours” went viral. Hits on the blog were around a hundredfold higher than usual. My guess is that people all over the world were alarmed by Kristi Noem’s press conference. But information about it was hard to get, because the media ignored it. So people searched and doom-scrolled. The heavy hits on the blog came from DuckDuckGo, Bing, Yahoo search, and Google.
I’m not entirely convinced that the problem is completely solved. I’d appreciate it if you’d leave a comment if you’ve tried to connect to the blog and encountered a problem.
The current controversy in Sioux Falls, SD, revolves around our city council unanimously voting to approve rezoning 164 acres of land so Gemini Data Center can build a data center here; a grassroots group is working to get engough signatures to put the matter to a public vote. The crux is that nobody wants a data center in their backyard, but everybody wants to post AI dancing baby videos on social media. It’s not hard to envision a world where Homepods are commonplace. As a teacher, I’m using AI more and more for lesson planning, text exemplars, and comprehension activities. Maybe Homepods would have a place in a classroom (?) As the demand for AI resources continues to grow, it’s going to be interesting to see where the resources come from and where the data centers go.
Hi Eric: We’re having the same data center controversy here. The county commissioners voted four days ago to rezone 1,800 acres for a data center, even though the citizenry opposed it and were darn near rioting at the meeting.
I can’t help but wonder if much of this data center building is speculative. For the proposed data center here, they apparently have not yet signed an occupant. It’s all being done with borrowed money, of course. I wonder what would happen if the data center bubble burst during construction and the money dried up.
I have experimented with locally hosted AI’s. They are not of course as good as ChatGPT, Claude, etc., but they will get better. Macs these days have lots of graphics processors, so AI’s should run really well locally on a high-end Mac.