I finally got down to Kmart today to pickup the microwave shelf that I saw online. I couldn’t find it on the shelf but lucky enough the guy at the local Kmart was able to locate it because it wasn’t on the shelf where it should be. I got home, assembled it and then organised it all:

It looks a whole lot less chaotic now that I’ve got it all organised which leaves more space for my kettle and toaster. While was was down there I picked up some food from Woolworths for dinner with mum on Tuesday – mum is bringing over the home made pumpkin soup and I’ve got the fresh bread bought today for tomorrow where I’ll add some garlic butter.
The artificial plant is on its way along with the woven planter that it’ll sit onside of – it’ll add some greenery to the space without me having to worry about taking care of it. The Ikea order is probably going to arrive in the next couple of weeks but I’m unsure whether the deliver date is the latest that it’ll arrive or whether it’ll arrive on that day. I’ve put a note on the order itself to leave the delivery on my back door so worse case scenario if I am working during the day I don’t end up missing the deliver then having to pick it up from the courier post depot.
I’ve finally folded up the fitted sheets and they’re in the linen closet along with the new sheets for the king single in the office. Once the Ikea deliver arrives I’m going to wash all the duvets separately along with airing out the new duvet inner that I bought. Although I have a queens sized bed I’d prefer to have a king size duvet on it which makes it all the more cosy especially during winter. The duvet inner is a king size but I opted for a woollen filling because the one I currently have now is also has a woollen filling.
Just before I go, I’ve noticed some really weird behaviour recently with Microsoft Teams and Safari where I keep getting the following error message:

And the issue has only just started recently and another thing that occurred recently is when uploading images to WordPress that it’ll just keep saying it is uploading event though the image has already been uploaded. I’m unsure what has happened because I haven’t changed any setting on Safari and I’ve got both WordPress and Microsoft Teams on my allowed websites in uBlock Origin Lite – the issue has only just started appearing within the last week or so. The only other thing I had changed was I installed the Safari Technology Preview 246 but I doubt that would have impacted it given that both websites were working perfectly fine when I had it installed. I have been experiencing bugginess with YouTube with pages not loading when pressing the back button but I’ve come to terms with my disappointment regarding Google and how broken YouTube and YouTube Music is.
With the rumours regarding a MacBook Pro Ultra, Microsoft announcing the NVIDIA RTX Spark – I’m wondering whether we’re seeing two big players hedging their bets that the future of LLMs won’t be in the cloud but being run locally. I say that because we’re already seeing companies after promoting ‘tokenmaxxing’ are now pumping the breaks. Uber pulling back on AI has given social permission to other b businesses to come out and admit that the promises made by the big LLM providers like OpenAI and Anthropic haven’t materialised. As Ed Zitron has noted in numerous interviews, the numbers for OpenAI are a pigs breakfast and I’d hazard to guess that Anthropic isn’t looking much better either.
As I keep reminding people about what a Sun Microsystems executive said over 20 years ago – the world runs on GET, good enough technology, which was in reference to the rise of x86-64 challenging traditional RISC based UNIX vendors and Sun Microsystems return to supporting Solaris for x86-64. This reality can be applied to LLMs – does it make sense to pay a premium to an LLM cloud provider or are you better off getting a machine with some good specs and run a local mode which is good enough to do the job without having the on going worry about how many tokens you’re spending? I think long term Google will win because there is a place where LLMs in the cloud makes sense but they also understand that developers will probably prefer not having to constant worry about billing that they cannot forecast based on the work they’re going to do. The big question is, does OpenAI and Anthropic have a plan B or are they ride or die with LLMs in the cloud.


