Welcome, to AI for Business Newsletter.
5 Answers to 4 Topics in 4 min:
1/ Detecting AI
2/ Using AI
3/ Privacy Concerns
4/ The Future of AI
Let’s get into it...
The Unique Twist (GPTs)
Integral part of this newsletter are AI for Business GPTs.
As you explore my content, the GPTs are offering personalized insights tailored to your needs and relevant to your circumstances. They align with your specific interests, correspond to your knowledge level, and adapt to your industry.
Behind the scenes, I am continuously training the GPTs based on practical knowledge I acquire from applying generative AI in my career and in business.
ChatGPT Plus account? Meet My AI for Business GPTs here.
Or try the free version below:
I have been talking to a lot of professional about Generative AI. In these conversations, a few key questions and themes keep coming up over and over again, influenced more by sensational headlines than actual AI capabilities. To clarify these misconceptions, I'm offering a FAQ based on these repeated questions.
I'm baffled how companies behind Large Language Models provide minimal user guidance, unlike even the simplest products like IKEA furniture.
Here, I'll tackle the most pressing AI questions I encounter, though bear in mind: I make mistakes, AI is shifting fast, and I may either be wrong, or will soon be wrong on this.
1/ Detecting AI
"Can AI writing be detected?"
The short answer: No. While there are tools claiming to spot AI-generated text, their effectiveness is questionable at least. These detectors often misidentify non-AI writing as AI, especially for non-native English speakers, leading to false accusations. Even OpenAI says detectors don’t work.
Your personal detection abilities: You might believe you're good at identifying AI writing, but in reality, this skill often only extends to recognizing poorly crafted AI text. Human biases and assumptions about AI usage also play a significant role. The truth is, after a few well crafted prompts, AI-generated content doesn't necessarily sound artificial at all.
2/ Using AI
"Where can I learn how to best use AI at work?"
Frankly, probably nowhere.
The bad news: there's no instruction manual out there to leverage AI in your specific field, and this is a challenge in navigating its benefits and pitfalls.
The good news: by using it a lot, you can figure out the best ways - it's an unique opportunity to discover methods tailored to your needs, and uncover valuable secrets in your industry.
"What is the best way to get good at using AI?"
To start, focus on the top three Large Language Models:
1. OpenAI's GPT-4: Currently leading the pack, accessible via a Plus subscription or through Microsoft Bing's creative mode.
2. Anthropic's Claude 2: A strong contender, especially for longer documents.
3. Google’s Bard: Trailing but watch for coming updates.
Finally, the best way to learn is by doing. Use AI wherever feasible - from brainstorming to summarizing meeting transcripts. About 10 hours of hands-on experience should provide a basic understanding of how AI can assist you.
Need ideas?
Ask the AI. Summarize a document.
Let AI assist. Draft an email.
Collaborate with AI: How can you help me in my job as [your role].
If you are new to AI, you may find my "AI for Business Tutor GPT" useful (it is built around an education context, and people have told me it was broadly helpful). Try it out here.
3/ Policy and Privacy Concerns
"My company won’t let me use AI because they don’t want the data stolen, is that right?"
Many companies hesitate to integrate AI over data privacy concerns. While it's true that Large Language Models (LLMs) may use copyrighted or biased data in their training, the actual risk to data privacy might be less significant than perceived.
For privacy, non commercial users can activate privacy modes that prevent data retention.
The answer is no. Organizations have even broader options than commercial users for ensuring data privacy. Many AI providers offer data-compliant versions and are willing to negotiate data guarantees to collaborate with businesses.
4/ The Future of AI
"What's next in AI?"
That's subject of debate even among experts. We're witnessing improvements as models grow larger, demanding more data, computing power, and financial investment. However, there are potential limits – technical, economic, or regulatory – that might eventually slow this pace.
Three scenarios are possible:
1. The development of Artificial General Intelligence (AGI), surpassing human intelligence and potentially transforming or challenging human roles.
2. Significant enhancements in existing AI capabilities, achieving higher levels of performance in various tasks.
3. A plateau in AI advancements, where progress slows significantly.
No clear consensus: the lack of consensus among AI experts means all these scenarios are viable. This uncertainty underlines the importance of considering each one.
One thing is clear: The AI you use today is the worst AI you will ever use!
Get insights between issues
To get more insights, follow me on LinkedIn, on X or let's talk AI.
Have a great day and until next time.
Roland
P.S. To sign up or share AI for Business Newsletter click here.
❝
Quote of the newsletter:
Comments