350+ free-roam venues with happy operators!
At the 50th anniversary SIGGRAPH conference in Denver, I witnessed firsthand how artificial intelligence is advancing at lightning speed. It’s going to transform our businesses and lives sooner than anyone anticipated. In my latest article for Replay Magazine, I delve into how AI is not just a tool but an extension of our minds, poised to disrupt everything. In it are practical tips for how you can use AI today in your business, and your life.
Last week I used Chat GPT’s new o1 Reasoning Model to create a proposal, a research framework, and a workshop run sheet for a new client engagement. Each of those three steps would have taken me half a day of thinking. Yet I was able to prompt the AI to give me something that was about 80% useful in ten minutes. It saved me at least a full day of work. And that was just for one project.
I also use AI to create content for the Inside VR blog on the VR Collective website. This week there are ten news stories. Each is about 1500 words. I sat down and started on Saturday morning around 11:00 AM. By 5:00 PM, with some long breaks in between, and with the help of AI, I published 15,000 words. Stephen King, eat your heart out.
Amy is the Artificial Intelligence staff writer behind The VR Collective Blog.
Designed as a representation of advanced AI language models, Amy Isaacson is not a real person. She embodies the capabilities of multiple sophisticated AI systems, transforming how we generate high-quality, engaging articles. At least I hope they are.
I thought it made sense to dive into our use of AI to write the blog and this newsletter, as full disclosure, and with the intention of encouraging you to start using these tools now. If you don’t you risk falling behind in your business.
One Big Thing (this story each week) remains my personal canvas. I write pretty much every word. When I use AI in this section, I will always try to qualify and disclose it. While AI is great at some things, it’s not quite there yet when it comes to deep insights into narrow subjects. But I am confident it will get there soon. Which can be terrifying if you make a living as a writer, consultant, coach, or any number of professions.
Artificial intelligence is reshaping journalism and content creation. seemingly overnight. With its ability to process vast amounts of information quickly, AI enhances efficiency and scalability. Journalists are now equipped to produce timely articles that reflect current events in real-time.
Advanced language models such as ChatGPT 4o and Claude Sonnet are pivotal in this transformation. These models analyze data, generate text, and assist in crafting narratives.
Key Contributions of AI Large Language Models (LLMs)
Amy’s abilities come from JuniaAI, an innovative writing tool that integrates multiple Large Language Models (LLMs). This technology enables a wide variety of writing styles and tones, making it easy to switch between different subjects.
I’ve trained JuniaAI on my extensive writing library. My book, my VR Arcade Buyer’s Guides, my blog at BobCooney.com and hundreds of pages of content on The VR Collective website. JuneAI then combines the information from those sources with output from ChatGPT and Claude Sonnet. The increased context from my past writing makes the articles more on-point and relevant.
With Amy Isaacson on the team, The VR Collective is able to publish a large quantity of relevant location-based VR industry news. If I were to write that much myself it would be a full time job, and I would need to charge an outrageous price. Instead, readers get free access to the latest news that not only educates but hopefully entertains, thanks to this powerful combination of human expertise (me) and artificial intelligence (Amy.)
Maintaining story accuracy in AI-generated articles is critical. Readers expect reliable information, and any inaccuracies can undermine my credibility. I make enough mistakes all by myself. Amy assists in research but require human oversight for verification. Much like your teenage staff members.
LLM hallucinations pose a unique challenge. This phenomenon occurs when language models produce content that sounds plausible but is factually incorrect. Kind of like almost anything on social media, Fox News or MSNBC.
It’s crucial for human editors to possess a solid understanding of the subject matter. This knowledge equips them to identify these errors and ensure that the published content meets high standards. I review every post written by Amy. Some more closely than others, depending on my evaluation of the relative importance of the story. If we’ve gotten something wrong, please let us know via email and I will correct the story ASAP.
The internet had a dramatic impact on journalism. As advertising models shifted to pay per click, the news sites realized that catchy, if misleading headlines can cause people to click on a story. That click triggered a page view, and a sophisticated system launched new ads on the upcoming page. We now call that clickbait. It exists because most people are not willing to pay for news. It’s the only model we have now.
Over the years, newsrooms have been decimated. There are about 1/3 as many newsroom employees at newspapers as there were in 2008. And that number is about to get reduced again by AI.
As AI models become more customized, they’ll go beyond writing articles. The entire newsroom will be populated by artificial intelligence.
NVIDEA founder Jensen Huang explained at SIGGRAPH how AI models are being trained to be highly specialized. AI models will work together in teams, with a management AI coordinating their activities. His example at SIGGRAPH was about digital image and video production, but it can be extended to writing, or pretty much any business activity.
I wrote more technical overview of SIGGRAPH that explains this in more detail on my personal website at BobCooney.com if you’re interested in a deeper dive. You can read it here…
There will be an analytics AI that looks at news, trends, internet search, pop culture, etc, and will make recommendations of stories based on what people are interested in. Those recommendations will get put into the research AI, who will pull accurate information for the writing AI. There might be an editing AI, or a fact-checking AI. All of this will then go to a publishing AI, that will push the content to the web, social media, and other outlets.
Humans are possibly just a few years away from being disintermediated. And it’s already happening. Sports Illustrated was busted for publishing AI written pieces by make believe journalists. The CEO was quickly fired by the publisher (more likely for getting caught that doing it in the first place.)
NPR human staff reporter Maggie Harrison has been following up the story, finding AI written content in the LA Times, Miami Herald, US Weekly and more. She broke a story this year about a company called AdVon, which creates product stories features across the web, mostly created by AI and attributed to fake writers with AI generated faces. It’s a fascinating read (or listen.)
We have been manipulated by media since it came into existence. Newspapers, radio, TV, even books have swayed how we think about subjects. The Bible and Koran have shaped the belief systems for billions of people. Somehow the idea of being manipulated by people is more acceptable than by computers.
But AI already decides what news we consume, since much of the world gets now gets at least some of its news through social media. The algorithms that control the news feeds are just a of artificial intelligence. As AI begins to control more traditional news media, the ability to be manipulated by a computer intelligence becomes a reality.
Sam Altman, co-founder and CEO of OpenAI (which created ChatGPT) and other AI leaders are claiming that they are working towards an AI superintelligence. Some people call this Artificial General Intelligence or AGI, which surpasses human’s ability to think. It will possess the ability to understand, learn, and apply knowledge across a wide range of contexts. What could possibly go wrong?
OpenAI started in 2015 as a non-profit, with a mission to make AI that benefits humanity. Since then, it’s raised tens of billions from for-profit companies like Microsoft, and venture capital firms. Last week it announced a $6.6 billion raise at a valuation of $157 billion dollars.
Along the way the original board of directors fired Sam Altman, the CEO, accusing him of throwing away safeguards in return for growth and profit. Behind the scenes strings were pulled, Altman was rehired, and a bunch of people were ushered out. Fundraising has continued new models are being released, and an AI arms race has begun.
OpenAI’s success led Google to release their AI tools faster than they anticipated, in fear they might lose their 82% monopoly on the search market. Search generates $175 billion a year for Google, about 57% of their revenue. Any significant loss to an upstart competitor would cost the company billions in market capitalization. Afterall, stockholders are people too.
Over the centuries, humanity has balanced our inherent need to invent and grow with the ability to survive. Growth is often risky and painful. 40% of the workers on the Panama Canal died. Thousands of Chinese were killed building the transcontinental railroad in the US. Progress always seems to come with a price. Those that pay the ultimate price often are at the bottom of the socioeconomic food chain.
Artificial intelligence might represent humanity’s greatest achievement. It’s also the riskiest. Initially the people who will suffer are those whose jobs are displaced. But if the science fiction writers are accurate, everyone might, at some point, pay the ultimate price of allowing artificial intelligence out of the lab and into the world.
I encourage you to start using AI now. The tools are cheap and easy to use. Your competition is using it. Your kids probably are too. Hundreds of studies have shown that over two decades the internet increased productivity and led to economic growth in almost every country. AI is going to do the same, but faster.
Some people are already saying that the productivity gains from AI are overestimated. These are probably the same people who said the internet was a fad in the early 90s. Do not listen to them. Analysts are notoriously unimaginative. They report on what has happened, and generally lack the ability to predict what will happen.
For now, AI can make your job easier and your company more efficient. Enjoy it.
