That is the setting. We can try to set the premise of DeepSeek (a waste of my time), we can set the premise of Microsoft AI (a waste of everyones time) and yes the 14 billion will have an effect and we can speculate on the 500 billion that StarGate is going to cost and what exactly will be the enabling part. Did anyone consider the ROI of that idea? That prospect will need to make at least 15 billion annual to make it worth. Throwing big printed cash at it will be as useless as the quantitive easing that Mario Draghi promised about a decade ago. Yup, it won’t go anywhere.
But that led me to a setting many seem to ignore, so lets have the list:
Microsoft 365 Copilot: A monthly subscription that costs $30 per person. Copilot Free is available with the Microsoft 365 Business Basic plan. Copilot Pro is a monthly subscription that offers more advanced features.
So at present, how many people are on this plan? It seems that Microsoft isn’t to talkative on ‘how successful’ it actually is. We get spread numbers and these numbers doesn’t seem to validate the billions invested.
Azure Machine Learning: A pay-as-you-go service with pricing based on the number of vCPUs.
Azure AI Search: A service with pricing based on the number of text records or images processed.
Here I have more issues. You see, we are given “Azure AI. Azure AI provides users with powerful tools that can be used to create innovative solutions using machine learning, natural language processing (NLP), computer vision, and more” How can any machine learning create innovative solutions? If it is machine learning someone else has it already, making it reengineering at best, optionally an innovative patent. I always (perhaps incorrectly) see pay-as-you-go as a dodgy solution. You either commit, or you don’t.
Computer Vision API: A service with pricing based on the number of transactions processed.
So, a service based on transaction processing, on that case if the IT department doesn’t throttle its usage there is every chance that an intern could blow up cost as it is happening.
Azure AI Content Safety: A service with pricing based on the number of text records or images processed.
Azure AI Content Understanding: A service with pricing based on the number of hours of content processed.
All this is set to a counter (like ConfirmIT) and that is the only company that had a good handle on it, a setting with decades. Now, there is a chance that I forgot a few solutions and that is OK. I am not heading an aspirational setting of academic instance.
You see everyone is on the bandwagon and I am too tired (or too old) to care. The media can’t be bothered unless digital currency is flowing their way. Yet in all this when did you see a clear description of AI solutions in use by Amazon, IBM or Oracle? You see, the DeepSeek issues of the last few days stirred a few minds. They are now also seeking Return on Investment (ROI) and that image is not clear, at least the media seemingly can’t be bothered and the influencers now shouting their wisdom on LinkedIn are also at times tedious and for the most a waste of everyones time. So why Microsoft? I don’t really care about it, but they (and their sickofans) are shouting how good their solutions are, but we see no clear numbers. And at present clear numbers is what the most of the population want.
Am I wrong?
I doubt it, the signs are there and when we see a small message on the left, the right clearly muffle that sound out. You see Shelly Palmer in IEEE Spectrum writes “As for the 100,000 jobs the project is supposed to create? Some construction jobs will be created as the data centers are built, but many more (millions more) will be created as the data centers come online. We’ve never had a compute cloud like this—there’s literally no way to calculate the economic impact of this amount of AI compute. It will be massive.” I actually don’t know about that. The idea that “there’s literally no way to calculate the economic impact of this amount of AI compute” is as I see it bogus. For 500 billion ($500,000,000,000) I expect more. But at present it comes across like a huge NSA data collection hub. Come to think of it, We could (optionally) get some data from the NSA, Google or IBM. They have experience with really big data centers. So what are those costs? What is the return on investment? And there is the setting of the value of collected data and that will not even have value until lots of data is collected, so lets say by 2030 and all those billions need to show investment value and at present the big-tech market lost over 1 trillion dollars a few days ago. So where is the ROI of all this?
Then we get “There are many tech skeptics, and it has become fashionable to denigrate and vilify big tech. To me, the Stargate Project is the first step in securing the future of the U.S. economy as well as our digital and cyber security. Every business will benefit from the power and promise of AI, and—like it or not, believe it or not—warfare will be dominated by AI. Today, the U.S. has a clear lead. The Stargate Project will help ensure it stays that way.” My issue is that there are always skeptics, I am one to some extent and the words “the power and promise of AI” fills me with dread. It is the included word “promise” and warfare isn’t dominated by AI, the setting pf properly programmed deer machine learning is. It is not AI and it is unlikely to show until somewhere in early 2040 at best (as I personally see it) but the 500 billion is coming out of ‘our’ pockets now. Yes, I know what they say that corporations will push the bill. Yet when this goes pear shaped. They will al put in in a bad bank account and relinquish the debt as a write off, so you, in the end still pay the bill in some way.
Then there is the sentence “Today, the U.S. has a clear lead” do they? DeepSeek is Chinese and their setting blew the rest away, you want to find out what a two-nil for China looks like? You are about to see that in very unrespectful terms. And as everyone is on that so called AI horse no one is investigating it, the media least of all.
In the meantime I will reengineer games. There is at least some revenue in that. And as I saw the reengineering options for ‘Infamous: Second Son’ The Sony firms could get some more coins from an 11 year old game on the PS4. And now there is an option to get it upgraded to PS5. Consider the gaming population. Whomever played in to PS4 (early days PS4) would like the setting on PS5, I tried that original game on PS5 and it plays well. A few minor glitches but that is what happens. The storyline could be upgraded and with linearity removed the game would get a much tougher stance. Then add the ‘cleaning’ of Seattle and we get a more complete game. With the setting to an optional change to Smoke-TV-Neon sequence the game alters a fair bit, and in this the game could also encase the stealth option in the game. Take with that the option to go back to the beginning to free the people from concrete affliction the good and the bad will also alter to some degree and it isn’t merely the good and the bad setting, the larger stage of animosity could reverberate through the game. And I am now looking to a few more games. A setting that I believe is great for Sony in the immediate future.
Can’t stop a creative mind puzzling on how to make something better, a trick that isn’t possible with Deeper Machine Learning and LLM’s. Have a great Thursday.
Lawrence van Rijn - Law Lord to be 