ChatGPT Unveils Multi-PDF Chat Feature, Is This the End for Wrapper Services?

ChatGPT just rolled out the ability to upload and chat with multiple PDFs at once. Yep, you heard that right—multiple PDFs. And while there are services out there doing something similar, they're often just "wrappers," and we've all seen how those are faring. Just ask Jasper, they learned the hard way when ChatGPT burst onto the scene.

So why did I step back from developing "wrapper" apps? One word: Microsoft. With their recently released Autogen, an autonomous agent program, and their proven capacity to handle a billion tokens in one go, it’s a signal of the seismic shifts coming our way.

In just 18 short months, we’re going to see AI models that are not just two or three times but hundreds of times more powerful than what we have today. Trust me, I've been writing about these advancements daily, and the future is looking nothing short of spectacular.

Switching gears to consultancy might not have the glitz and glamour of app development, but it’s where the real need is right now. There's something incredibly satisfying about watching someone’s jaw drop when they realize that a task that took hours now takes mere minutes, thanks to AI.

Just last week, I heard, "You've saved me an entire week of headaches in a matter of minutes." And it never gets old.

Currently, I’m knee-deep in discussions with a few industry giants about AI integration into their existing frameworks, and simultaneously assisting some kick-ass small companies to get their teams up to speed with tools like ChatGPT.

As always, I'm grateful to everyone who’s been tuning into the newsletters. With nearly 200 editions and a 49% open rate, we’re definitely onto something big here. But hey, I still need your support. I’m setting my sights on 10-15 new client partnerships in 2024 as a consultant/trainer and even as a fractional Chief of AI Officer. Plus, I’m also open for more podcast invitations and speaking gigs. Every mention, tag, and introduction counts, and I can't thank you enough for them.

Robo-Advisors Still Have a Ways to Go Before Replacing Humans

Financial advice bots like ChatGPT are all the rage these days. Everyone's wondering - can AI really replace human financial advisors? The short answer: not yet, but the bots are getting close.

Here's the deal. Good financial advice requires five key skills:

1. Debiasing - helping clients avoid dumb mistakes caused by normal human biases. Bad news here - the bots are super biased! Just like us. For example, they want to "make up" losses by taking more risk, rather than cutting them. To fix this, we need to program in some "meta rules" that force the bots to second-guess their gut instincts.

2. Empathy - reassuring anxious clients so they don't freak out and sell. Surprisingly, the bots kick butt at this! In tests, healthcare pros said ChatGPT responses were way more empathetic than real doctors' answers. Bots could be a game changer for helping nervous investors stay the course.

3. Accuracy - getting the facts right. More bad news...the bots make a ton of mistakes right now. Ask them to compare two funds and they'll often pick the wrong ones! This is an easy fix though. Just link the AI to legit financial databases to check its work.

4. Best interest - avoiding conflicts of interest. No evidence bots are better or worse here yet. Their advice will need auditing, just like a human advisor. But monitoring AI is likely easier than monitoring a human.

5. Consistency - giving similar advice to similar clients. Studies show human advisors struggle here. The good news is AI should kick butt at consistency once trained right. It's like Netflix recommending shows based on your tastes.

The future ideal model may be a hybrid human-AI team, like a doctor using AI diagnostic tools. The human filters and interprets the AI's suggestions. This will likely make the advice way better.

Even for folks who can't afford a human advisor, AI could still work if we build in the right safeguards on bias, accuracy, etc.

So if you're an advisor, don't stress about the bots taking your job just yet. Focus on using AI to give more people access to quality advice! With the right human oversight, these tools could make financial advice way more effective for everyone.

AI Skills Now the Golden Ticket to 40% Higher Pay, According to Landmark Study

Alright, here's the deal folks: Want to earn a fat paycheck? Get AI-savvy. A new study is shouting from the rooftops that if you've got AI chops, you could be raking in up to 40% more than the Joe Schmoes who don't. That's not chump change, people.

This Oxford Internet Institute and University of Copenhagen study isn't some dinky online quiz; it looked at 25,000 freelance workers. And guess what skills employers are drooling over? You guessed it: machine learning, open-source AI software, and deep learning. Dr. Fabian Stephany, one of the brainiacs behind the study, is practically handing out a roadmap for your career. He's like, "Hey, we're in a tech revolution, better reskill or get left in the dust."

Look, this isn't a fad. Last year's ChatGPT launch shook the corporate world. It's so critical, Unilever's CTO Adam Raeburn-Jones even weighed in at Google Cloud Next. According to him, if you know your way around AI, you're essentially the LeBron James of the job market.

But here's the kicker: Stephan Pretorius, CTO of ad agency WPP, says the young blood entering the workforce now are not just tech-savvy, they're tech-native. Forget the learning curve; these folks are starting on a whole different plane. And you know what? They aren't naïve. They're eyeing companies that will let them flex their AI muscles.

Columbia University's Oded Netzer says the focus on AI skills isn't limited to Silicon Valley. From coding to marketing, if you don't catch this train, you won't be replaced by robots but by someone who knows how to boss them around.

Key Takeaways

- AI skills can hike up your salary by as much as 40%.

- Young workers are entering the job market with a fundamentally different skill set, focused on AI.

- If you're not AI-savvy, you're not just at risk of stagnating; you risk being completely overshadowed.

AI in Healthcare: From Personalized Doctor Recommendations to Predicting Mortality—The Promise and Pitfalls

Artificial intelligence is making waves in healthcare, and not just in theoretical papers or controlled trials; it's showing up in practical ways that could soon touch all of us. Let's look at a few developments in AI that stand out for their potential impact on healthcare outcomes.

Imagine AI systems, like ChatGPT or Google’s Bard, pulling together data from various sources to give you personalized recommendations on which physicians and hospitals have the best track record for your specific condition. It's like Yelp, but for your health. However, the quality of this information is contingent on how reliable the underlying data is, so transparency is key.

Now, what if I told you your voice could reveal you have diabetes? That’s right—recent studies indicate that AI can analyze subtle changes in voice to identify type 2 diabetes. With accuracies of 86% for men and 89% for women, the findings hold promise, although further validation is needed.

Switching gears, researchers from Vanderbilt University and University of Missouri–Kansas City have discovered another practical application for AI in healthcare. By analyzing CT scans originally used to screen for lung cancer, the AI model could also predict mortality from cardiovascular disease. This could be a game-changer for preventive medicine, enabling interventions even before symptoms appear.

However, the use of AI in healthcare isn't without its challenges, especially ethical ones. Predicting a patient’s mortality within a short timeframe could lead to complex dilemmas in emergency rooms or hospice settings. While professionals see AI as a promising tool for better healthcare, they caution against employing it merely as a cost-saving measure.

AI’s growing role in healthcare is promising but also prompts ethical considerations that can't be overlooked. What’s clear is that AI can serve as a potent tool for individualizing healthcare, aiding in early detection, and potentially saving lives. But as we integrate these advanced systems, it's crucial to handle them responsibly, ensuring they augment rather than compromise the quality of healthcare.