Sam Altman wants to build Deep Thought
PiP shorts #5
Thinking about AI, a cost benefit analysis
OpenAI Boss Sam Altman Seeks *Trillions* to Fund AI Chip Production (breitbart.com)
There is just one thought I want you to contemplate:
This is not a charity, but a venture capital fundraising. Investors are supposed to get a return on it, and since that return is quite a few years away, it has to be a large multiple (at least 10x) of the initial investment.
One must wonder: where will this money come from? Ultimately, it will have to come from us, the consumers of whatever benefits and services it will provide.
Sam Altman is building LLMs (Large Language Models), not narrow AIs, that are task specific tools.
His (Altman’s) aim is to create Artificial GENERAL Intelligence, that is ‘smarter’ than us.
The return will come from the companies and entities who will be profiting from manipulating us.
Just like it is the case with search engines and social media advertising. LLMs are just the beefed-up versions of the search and social media algorithms.
You don’t have to go far to see where this is taking us:
Sam Altman is a Davos Man. Elon Musk is suing him for the betrayal of the company’s founding principles.
Ultimately, the profit, the return on that 7 trillion-dollar investment will come from us. We will pay for our own brainwashing. We will assist in our own enslavement to the investors of this grand globalist enterprise.
If the goal is to build a computer that can design a better mouse-trap, by all means, invest in it, but there is absolutely no reason to invest in Deep Thought, as we already know the answer. It is 42.
Somebody should tell Sam Altman…….
References:
The TED Interview: The race to build AI that benefits humanity with Sam Altman (from April 2021) | TED Talk
Elon Musk Sues OpenAI and CEO Sam Altman for Failing to Use AI for the Benefit of Humanity
Davos 2024: Sam Altman on the future of AI | World Economic Forum (weforum.org)
(This is a scary stupid conversation)
AI's Impact on Humanity: Sam Altman's Davos Speech — Eightify
The Impact of AI: Insights from Bill Gates & Sam Altman — Eightify
At the end of this post:
….you will find links to a series of my posts on AI.
Just after I scheduled the above for posting, I came across this video, confirming my points about the problems of the LLM strategy with the results of empirical research:
There must be a more complete context of Artificial Intelligence than "I AM THE FUTURE".
There is the human fascination with the phraseology, "smarter than we are". The phrase has no given meaning whatsoever. The outside trappings are the ability to survey mountains of data in an instant, sort it out for the trends that it may contain, select those identified trends which conform to "politically correct", (or whatever the programmer's built-in bias is), and quickly spit out some answer that has a modicum of credibility.
Is that what is "smarter than us"?
We all know that data is the profit engine of social media. We all know that data is past events and projections of what we will do based on those past events, (trends). We all know that data collection (with your identity), is the primary function of cell phones, smart appliances, "smart cities", the operating systems of Apple, Microsoft and Google, (iOS, Windows, and Android), and all the tracking and ring cameras, facial and voice recognition., Iris scans, and all the social credit systems now in practice.
Mr Sam Altman dreams of ONE-UPPING all those other idiots just mentioned above, and build a system with so much credibility that you will voluntarily read-into it your every thought, intention, and action.
DEAR GEMINI: Please tell me, what is my shopping list for today? (Well, he'll get it right.)
Surpassing the current surveillance model, by many orders of magnitude is the dream. Billions of people standing in line to get into the commissar's office: "Here is my self-report for the day". Hey; this office is crowded, you can please go make your report on-line.
But be sure, achieving this dream is of the least importance.
Any time vast sums of money are moved around, vast fortunes are compounded every day. It is only when everything is deemed OK as it is, and that with no movement - nothing happens. The old favorite is the military, because an enemies-list can be created at the drop of the hat, and with security issues, nothing can be audited, or even spoken about. (It doesn't even exist.) This will always be used to the fullest. But there has to be some ADD-ONS while the military is fully exploited.
Okay, Global Warming is going to extinguish life on earth. It is all energy's fault. Greenhouse gas. (By the way - fossil fuel built civilization.) Before coal we burned wood and dung. That life lasted many 1,000's of years, and could just as easily still be going on today.
In 2022 the world burned 180,000 Terra Watts of energy, (who cares what that means), but 16% could be called some kind of renewable. The greatest feature about renewables is that night time has no sun and maybe 60 nights per year have no wind. Also output changes with every passing cloud and puff of wind. So redundancy needs to be built into the system.
Furthermore; the distribution system CANNOT be subjected to shocks, and most forms of energy generation cannot be tuned in and out at will, to counterbalance the capriciousness of renewables, and still keep the loads stable. That means that the Green-New-Deal has to keep building way past the global use of energy, maybe to double or triple it. Of course all of that building adds immeasurably to the green-house gas that you are claiming to eliminate.
Nevertheless the Green-New-Deal is spelled B-O-N-A-N-Z-A.
Back to A.I., Sam Altman makes the calculation: If I am to make a Large Language Model, and if there are this many billion operations added every minute, I am going to have to access X-Gazillion units and sort them for my algorithms in a time that the enquirer will not get bored waiting. Of course I can't do that, but given the trend of chip size reduction and speed increase, and projecting the quantities of operations that I envision??? (It is grade school math, but with lots of zeros added onto it.) Well it will have to be in this amount of years, and I will (someone must) invest:
$7 Trillion! Why that's a good number for starters. (I would say that he asked Google Gemini?)
It makes no difference what fantasy numbers he comes up with, enough to hit the news, and once any money rolls, not only are fingers sticky, but money is sticky in itself, and is looking for something to adhere to. Just remember, that the building of A.I. is a "RACE", and that the Chinese will win, but we hope to come in a measly second.
I ran out of words to discredit A.I., but to say; a fervor generated by billionaire's patsies.
Of course I also have a great invention, which may require trillions to get going and it will truly save mankind. Instead of making a "remembering-machine",
I have the plans for: A FORGETTING MACHINE. Plug in this "fucker", and all of your problems whatsoever, will be solved.
.
While there is good reason to believe that there are risks from putting such power in the hands of some very powerful people, but there is also room for optimism if you believe that technology has been a net positive for humanity.
Some part of me hopes for a moral and ethical Deep Thought platform that could become an arbitor to try and convict government leaders and billionaire monopolists who start wars “for profit” or enslave humanity “by edict” through various control mechanisms. Such people can’t be trusted and history has proven time and time again. In an ideal scenario, an Deep Thought Advisor for Humanity could change the trajectory of history to one that can leverage the benefits of Capitalism to continue to bring benefits to humanity while stopping preditors from exploiting humanity for their own power and profit goals.
I prefer a balanced discussion about AI - a SWOT Analysis approach - that will help all of us to obtain a 360-degree understanding of AI. So far, I have found productive and helpful uses of ChatGPT while also understanding its limitations and built-in biases. For me, this experience is no different that many IT tools and systems I have used in the past.