Well, the week is finally over and the weekend has started – I was feeling a bit peckish so went down to the local store down the road to pick up a couple sandwiches and a packet of crisps. The weather is nice outside so I sat out on the back patio and had my meal – enjoying the fresh air and chilling out after having a rather hectic day. This week I’ve got a broken weekend where I’ve got Monday off then in the office on Tuesday and then Wednesday off from work- it isn’t ideal but everyone in the office is having to do it. Given how jobs are difficult to come by, the economy is looking precarious and there is an election, I feel happy that at least I’ve got a job even if I do regularly complain about it.
The new setup I’ve got in my office is going really well and on Monday I’ll be taking the monitor along with the old equipment into the office. I’ve got a $15 voucher in my ‘Everyday Rewards’ where I’ll need to pick up some ‘Up and Go’ for my breakfast and maybe some few other items. What I’ll do before I leave tomorrow is to put together a list just to avoid standing in the isle and asking myself “hmmm, do I need to purchase that? I don’t remember seeing it in the cupboard…” only to find that if I buy it I have a double up and if I don’t by it I get home and realise I didn’t have it after all.
I’ve been having a read through Reddit and it is always funny hearing people talking about rumours as if they were statements of facts – posters declaring that x product is the worse product that would ever exist and that they’re sticking to what they have now all while ignoring that said product hasn’t even been launched yet. I’m still doing some contemplation about what I’m going to do with things going forward because I have a soft spot for Android because the biggest strength over iOS is the fact that it isn’t a monolithic operating system meaning that the individual components can be upgraded outside of the big releases. When it comes to iOS we see an update every 6 weeks or so but that tends to primarily bug and security fixes where as with Android you have your monthly updates plus individual components such as Chrome as not just updated but upgraded with new features for both end users and developers.
What has the end result been of iOS not being able to upgrade and update on a modular way like Android is? have a look at the Web Platform Test website comparing the latest stable versions – Chrome is sitting at 93% compliance where as Safari is hovering down at around 80% compliance with extension support of Safari on macOS absolutely abysmal because Apple insists on implementing Declarative Net Request (DNR) on top of the Content Blocking API rather than implementing it natively in Webkit (which results in one having to wait for the DNR filter to convert to the native format) not to mention the bugs that exist, I follow on GitHub, I request a status regarding how far away the fix is and Apple developers ignore my posts. At this point I wonder whether Apple developers even care about making a decent product given how little there is a sense of urgency of addressing the product’s shortcomings.
I saw this video on YouTube from the BBC regarding Meta and Apple teaming up which raises many of the issues that I raised not too long ago:
We’ve started to see the market finally wake up and ask the obvious question – not whether the technology of LLMs (or AI in general) is useful but whether there is even a business case to answer for to justify the hundreds of billions being spent on this build out. At some point after billions of dollars are burnt we start to see people ask the question, “ok, now you’ve burnt billions when are we going to see a return on that capital expenditure”. This is was what eventually took down the whole dot-com bubble because big promises were made, a lot of leeway were given to these dot-com businesses but at some point I think it is fair that those providing the money as capital to start asking when does that investment turn into a viable business model which will in turn become a sustainable long term profit growth.
There is also the element of higher interest rates – with interest rates going up the gap between the amount one can get from a term deposit vs buying bonds or shares has to be sufficient enough to justify the risk one is taking. If people start seeing the interest rate go up, the gap between the interest rate on a safe investment vs a speculative investment close then I wouldn’t be surprised that fewer people are willing to throw money into the pile. For example, OpenAI (link) is struggling to hit its funding round goal as people are finally asking what the long term plan is other than burning money and now there is reports of Anthropic bonds being rated as junk following its IPO (link) by credit rating agencies. At some point the interest rates on bonds are going to get so high that even those with an appetite for risk will just throw in the towel and walk away because it is becoming ridiculous.
The alternative vision to this whole cloud based LLM build out is open weight models running on local hardware – either on the workstation or in the data centre of a company where it is tweaked and fine tuned to do specific tasks where LLM excel at. The reaction by Sam Altman to a future where he isn’t sitting at the entrance of the cloud clipping the ticket for every request into the cloud and ever response coming out the other end? he blames end users for wanting to have their own a computer for why RAM prices are going up and not because of the speculative bubble the he and his ilk have created off the back of fansifull narratives about a post scarcity future where we push beyond the limitations of our current productive forces (link). Just when you thought the broligarchy couldn’t get more out of touch we have Sam Altman showing that yes, they could get further out of touch.
