There can only be One

April 12, 2024

DENVER, Colo., Apr 11, 2024 (247marketnews.com)- Since the smart phone’s introduction, there has been a lack of noteworthy innovations. One could argue the creation of software is innovative; however, if you really think about it, it’s just a way of implementing the “smart” phone’s innovations into our daily lives. That’s one of the many reasons why most apps have come and gone, since the iPhone hit the market in 2007.

Connecting objects to the internet of things is often mistaken for innovation, but it’s usually just performance or maintenance analytics tracking, which may be handy, but far from transformative.

So called Artificial Intelligence, or AI, is simply Large Language Models (LLMs) and predictable human speech patterns, based on available data points to mimic human speech with a wide breadth of knowledge.

While not truly AI, this innovation confirms that technology is getting closer.

Yet, AI has been all the rage on Wall Street, over the past year, and merely mentioning AI implementation into existing operations has sent numerous stocks soaring, in a craze that’s reminiscent of the 1999 frenzy that preceded the dot com bubble bust.

AI revolution rumblings were already causing shockwaves before the Chat GPT launch, when a Google engineer claimed that their chatbot was sentient. These claims alerted us to prepare for the launch of the AI future we’ve been dreaming of since Star Trek and C3-PO, or fearing since the Terminator.

Powered by Nvidia’s (NASDAQ:NVDA) H100/A100 GPUs, which enable servers to simultaneously process larger amounts of data, in real time, allowing for an unprecedented approach to analytics, LLMs with a billion or more parameters can now be compared, viewed, and analyzed instantaneously. Way too much data for humans to interpret, but a snap for the new chips and newly created algos.

These chips and cards allowed software makers to create software for robotics, 3-d visualization, database query processing, medical data processing and seemingly endless uses. However, when it comes to AI and direct human interaction, how many generative AI models do we need and should we allow more informational user interfaces?

Data is only as good as the source and researchers recommend more than a quick cursory search for something, but what happens when Chat GPT or Bard provide incorrect responses?

The current leaders in this space are, as expected, have huge global user bases and large data sets for AI modeling, like (NASDAQ:MSFT), which is now partnered with OpenAI/Chat GPT, Elon Musk’s X/Gonk, Apple (NASDAQ:AAPL), Meta Platforms (NASDAQ:META), Alphabet (NASDAQ:GOOGL)

All of the data uploaded by humans, indexed by search engines, and published for online consumption are now part of the LLMs and will serve as the foundation for upcoming AI models.

Contemporary AI is a giant data scraper, with parameters set and algorithms created by various humans. It does not actually learn anything; it only does what the program instructs it to do based on previous results.

The consensus from a recent discussion with several long-time programmers was that Google clearly had more data about our thoughts, questions, and information. From the most mundane question to medical diagnosis, the search history and index of information required to run the more than 8 billion daily search queries would undoubtedly be the winner.

Microsoft’s CEO recently confirmed that consensus when he stated that Google “should have been the default winner” of Big Tech’s AI race, thanks to Google’s resources. “Google’s a very competent company and obviously they have both the talent and the compute. They’re the vertically integrated player in this.”

So, everyone, and I mean everyone, agrees that OpenAI should have been crushed by Google. In fact, CHAT GPT uses Googles own indexing to facilitate and run its LLM data sets.  It literally scrapes Google.

Google then released Gemini and broke the internet for the wrong reasons.

Sundar Pichai and Deep Mind CEO, Demis Hassabis, described the Gemni AI era as, “A Huge leap Forward in an A.I model that will affect practically all of Googles products.”

This statement was meant to show the world what should have been the most robust AI LLM in existence, but it was quickly overshadowed by what appeared to be “woke” programming.

Google’s AI became a laughing stock for creating images of African-American George Washington, African Nazis, and other abhorrent images.  The Image generator created female Popes and stated that, “It’s not appropriate to assume all popes are Caucasian, many have been of European Decent.”

The Gemni project release was a disaster for the company and had to quickly have the image generator shut down, to be reeducated, and Jack Krawczyk, Google’s AI product lead, faced backlash for the Gemin Inaccuracies and eventually removed himself from social media.

Gemini wasn’t the only AI LLM to be programed with this bias, but it was probably the most egregious, proving, without a doubt, that AI and LLM models must be reigned in, by a strict set of ground rules before they get out of hand.

Humanity deserves accurate information, so we can make the best informed decisions, and that is what AI should be all about. Anything else is a gimmick.

AI should be like CP30 from Star Wars and unemotionally relay facts and statistical outcomes for any given scenario. It should not care about color or have an opinion on the data, unless its relevant to the outcome. It should simply give the user what it needs to be successful for every task.

The current LLMs are better than the AOL chatbots that I helped create, long ago, for marketing on AOL messenger, but they still have a long way to go.

The current control models leave a lot to be desired from the current models being touted as AI. Who is programming the algorithms that determine the correct answer to a question? Who decides which authority ranks higher and do we give solid arguments where an issue has good arguments on both sides? Do we compare and contrast likely scenarios and outcomes based on math and real-world data?

There are currently as many AI apps as there were phone apps during the app heydays and there won’t be a shortage of companies proclaiming AI advancements and LLMs that will end up into the dust bin of history.

On top of this, advancements are being made to put all of humanity’s data into a chip that will be installed or connected to the human mind, enabling individuals to become super smart.

Data with human imagination could create super intelligence and, again, points to Musk, but this time for his Neurolink project.

OpenAI and Chat GPT should be the clear winners in this space, as the others are just riding the wave. Musk, Gates and Altman have the right concept, and will make this a very exciting space to cover for the foreseeable future.

About 24/7 Market News
As a pioneer in digital financial market media, 24/7 Market News (24/7MN) is dedicated to the swift distribution of financial market news and information. 24/7 MN incorporates comprehensive corporate communications resources and tools to engage the investment community. Visit 24/7 Market News website here.

24/7 MARKET NEWS, INC Disclaimer
Please go to https://247marketnews.com/disclaimer/ for disclaimer information.

CONTACT:
24/7 Market News
Editor@247marketnews.com

Related news for (NVDA, MSFT, AAPL, META, GOOGL)

NASDAQ and NYSE quotes and data are delayed 15 minutes unless indicated otherwise. Market data and exchange information are provided for informational purposes only and is not intended for trading purposes. Neither 24/7 Market News Editors, 247 Market News, or data and content providers shall be liable for any errors or omissions, delays, misquotes or other market information relayed in any press materials. You should Use Realtime data to conduct due diligence before investing or trading, and trading in any stock is risky you could lose all your money.