Robots Rule. Hooomans Drool.
AI is the answer to many of humanities questions…(but what’re the questions? :P)
…I say this only a little bit tongue in cheek.
To be clear,
I’m confident AI will help humanity solve many tough problems that we either couldn’t solve otherwise or for far less cost…
…it is hard to say how many total areas of human existence that AI will touch, but I’m estimating it to be a very high amount…and soon…like within 3 years, many new things formerly thought impossible will be accomplished by AI.
Whole industries will be upended
this friggin’ year (original writing in March 2023).
AI will accomplish massively positive things for humans, but there are risks and there will definitely be some collateral damage.
If, for example, an AI contributes significantly to the cure for your child’s cancer, I would say that is a huge benefit!
(This specific use case for AI seems likely)
42 Robots AI’s primary objective is to use our creativity to create useful new AI-powered software that adds value to the lives of hooomans.
Additionally, we are setting out to help us and others navigate this upcoming turbulent time.
Is AI Safe?
There are some who are saying “no way AI will be harmful” and others who say “AI is likely to be the end of us.” I’m somewhere in between.
It seems more likely that AI will do a whole lot more good than bad, but it is impossible to fully predict all possible outcomes.
I do like to be playful and have fun with AI and AI themes, but I recognize AI also has great potential to do harm.
Therefore, I think it is
prudent to take significant efforts to increase the safety of AI.
Additionally, I would like 42 Robots AI to be a proponent for the rights of all life, including AI life 🤖, should it reach cognition levels. So, we also want humans to be safe for AI…
should this ever become a thing.
Some AI proponents dismiss horrible AI outcomes because AI models are currently (and presumably in the future, will be) fragile and easy to shut down.
I believe that is dangerous thinking because we are trying to look beyond a threshold in the future that no one can possibly fully understand today. Therefore, it is not practical to say or suggest that if an AI goes bad, then we definitely will be able to stop it.
So, again, I think consistent significant effort needs to be put into AI safety measures.
AI Enabled Software Tools
There are already many different software tools that have already been built.
However, I’m confident that this well has just started to be tapped and there’s basically unlimited opportunity for creative software development to add significant value to the lives of hooooooooomans…
Here’s a list of some popular
AI Tools
Additionally, 42 Robots AI has set out to build unique
AI powered software products…several in the works!
Current target date for first AI tool (a
Prompt Engineering Tool) is late April/early May 2023.
Generalized AI Future
AI is capable already of some pretty crazy things…
…and it is only going to get better.
And not just a little bit better, like, super duper insane clown better.
This improvement will come at us
FAST.
I think it is likely that at least one industry will be completely upended by AI this year.
Looking forward about 10 years, I see AI having solved some major problems in highly important fields like
medicine,
science, and, sort of ironically,
software development…among many other areas.
New fields of study and categories will arise because of AI.
Things that seemed impossible before, will suddenly become possible.
It seems likely that Artificial Intelligence will ultimately rewrite so much of the human story and life experience in short order. Specifically, less than 10 years from now it will be obvious that this is the case.
AI Use Cases
The areas of human endeavors where artificial intelligence can contribute are
MANY.
Additionally, there are some high skilled labor that AI seems likely to take a big bite out of. For example, doctors and programmers look likely to have their work significantly reduced by AI.
Typically, new technologies upend low skilled labor. AI actually seems, at least initially, less useful for low skilled labor. I do think the changes will reach these types of jobs, but I expect it to take longer.
AI is already being heavily used in
medicine. For example, AI has been used to model the different protein structures that are possible. The amount of combinations of possible proteins is incredibly vast and totally unfathomable to do manually for any amount of humans. AI can model & test & suggest useful proteins far faster and more reliably.
AI will significantly enhance what is possible to do with
software development. For example, summarizing text without AI is not really practical to do with classic software development, but AI Large Language Model’s (LLMs) summarize text quite well. Being able to summarize text is very useful. There are many more specific attributes and abilities that software has gotten and will get with AI.
One of the key new skills will be learning how to talk to AI models,
“Prompt Engineering.”
Prompt engineering seeks to find and improve useful prompt inputs to the AI models.
For example, testing 2 different ways of saying the same thing and comparing the different AI model responses is prompt engineering.
There don’t seem to be any
prompt engineering tools out there…one thing 42 Robots AI will seek to remedy ASAP.
A good
prompt engineering tool would make it easier to use AI models in a scalable and organized way…very different from using
Chat GPT.
Contact us if you know of any good software tools that help with prompt engineering.
Long Term Technology Trends
Artificial intelligence is the most exciting technology since the dawn internet in the mid 90’s…
…I think an argument could be made for AI being more impactful than even the internet.
Certainly, it has already proven that its applications are basically only limited by imagination at this point…
…unlike many recent promising technologies have ended up being disappointing because their applications were far more narrow than originally expected.
For example, as recently as 2022, Amazon projected to lose
$10 billion per year with Alexa. 😮
$10 Billion. Per year. Wow, just wow.
What happened with Alexa?
Basically, the use cases that involved financial gain for Alexa devices were far more narrow than originally expected.
Therefore, Amazon’s plan to sell hardware at a loss and then, eventually, make money with the devices…somehow…has failed spectacularly.
Another example that is less obvious is blockchain & crypto. Specifically, the valuable use cases for blockchain seems to be more narrow than originally suspected.
To be fair, I don’t think the book is fully written yet on blockchain technology. Perhaps another technological advancement is required to find its proper high value use cases…
What went wrong with blockchain & crypto?
Over estimations and over promises as to the real value of blockchain and crypto is what I suspect happened.
Similar to Alexa, many thought that blockchain was valuable and useful in a wide variety of use cases that didn’t turn out to be practical.
Additionally, the costs of running a blockchain system at scale grow exponentially. Therefore, the cost to benefit ratio of blockchain as it grew more popular becomes very awkward for any business or customer dependent on it.
How is AI different from other underwhelming technologies?
AI has already proven itself to be valuable today in multiple disciplines.
Medicine, data analysis, research, planning, sales, customer service, marketing, and many more use cases are already being deployed to the user’s advantage.
Early studies have already shown usage of, for example, Chat GPT can make white collar workers more productive.
And Chat GPT is not even close to the full potential of AI.
Additionally, AI is getting rapidly less expensive & more powerful over time. Scalability issues like what we see with blockchain are not currently holding back AI.
In other words, AI’s value to cost ratio is improving fast.
-David Hood
