We Can't Trust Everything AI Tells Us. And Here's Why That's a Good Thing.

Hi friends,

For all its promise and hype, AI makes mistakes too.

Imagine AI as an endlessly enthusiastic but occasionally fumbling trainee, not a perfect genius. It undoubtedly holds the power to transform entire industries, but its path is paved with growing pains and stumbles that require our seasoned guidance.

AI is fallible, it can make mistakes, conjure up the odd, make stuff up, and harbor prejudices. But these aren't roadblocks; they're opportunities to put our knowledge, experience, and good judgment to work.

Simply by collaborating with AI, we're shaping this powerful tool by merging our human insights with AI's unrefined potential.

And that’s a good thing.

WDIM?: By viewing AI as a partner, not a panacea, we're not merely leveraging AI, we're growing alongside it, honing its intelligence, and ensuring its development aligns with our collective values.

AI SMORGASBORD

Why tacking Deep Fakes has turned into an arms race. Technology has developed to the point where human discernment will soon be impossible, and disinformation can be devastatingly effective. Is more technology the answer? 👀 

 “Talk to the hand” takes on a whole new meaning when you meet the Ring Whisperers. Watch out for the development of ever-more innovative, wearable communication interfaces between us and our AI Assistants 💍 🤖 

Adobe is ramping up its AI capability with its latest offering GenStudio, an “AI factory For Advertisers”. We’re seeing more platforms bringing more services under a single umbrella, as people clamor for less fragmentation and better end-to-end solutions. 🏭️ 

Power grab! We don’t see as much discussion about AI’s power-guzzling and carbon emissions as one would expect in such climate and energy-conscious times. Yes, how can the world strive to achieve net zero when we keep inventing technology that consumes vast amounts of energy? (On the flip side AI is being used to build smart grids for more efficient energy management) ⚡️ 

The World Bee Project is the first independent initiative in the world that uses AI and advanced technologies to monitor pollinator and biodiversity declines from a global perspective, to seek dynamic long-term solutions to benefit both nature and people. 🐝 🐝 🐝 

AI MADE EASY

Mastering design isn't required to craft stunning ebooks, reports, or any polished documentation with these three powerful tools. My one-day ebook creation challenge (to myself), while extended by tea breaks and tool exploration, ultimately took around six hours.🫖🍵 🍵 🍵 

WDIM? Sometimes we look for a single solution when it requires knitting together a few tools to do the job. Don’t be afraid to experiment!

E-BOOK WORKFLOW

AI TOOLKIT

Are AI Tax Assistants ready for prime time? While they may not yet be stand-alone, there is a ton of ongoing development in this ‘AI-ripe’ sector. From handy GPT baby bots that have consumed entire tax codes, dedicated AI platforms ready to break up the Turbo Tax and HR Block monopoly, and market leaders hurrying to incorporate AI into their platforms, AI tax help is at hand. WDIM?: AI is making tax preparation so much faster and easier, but it’s not only taxpayers who see the benefits. The IRS is using AI in customer service chatbots, and for auditing!

AI SKILLS-SHIFT

Singapore is not planning to be caught short by AI developments. Instead, they’re proactively preparing their citizens for the skills-shift ahead. They explain their plans in this short video, which is well worth a watch.

AI-FLUENT?

Are you comfortable talking-the-AI-talk? Boost your understanding and vocabulary with our plain English glossary.

WHAT IS AI BIAS?

In the context of AI, bias refers to a systematic error or unfair preference in the decisions, predictions, or behaviors of artificial intelligence systems.

Here’s an analogy: Imagine teaching a child to recognize what an engineer looks like by showing them a photo album. If this album is filled mostly with pictures of men wearing hard hats and holding blueprints, the child might start to believe that all engineers are men who wear hard hats. This is how AI develops bias.

AI, much like the child, learns from the examples it's given. If we feed it data (like photos or resumes) that only show a narrow slice of the world (for instance, engineers or resumes being predominantly of one gender) the AI starts to treat this narrow view as the 'norm.' So, when it's tasked with selecting potential job candidates or identifying engineers, it might unfairly favor those who fit the narrow profile it has learned, not because it wants to, but because that's what it thinks is correct based on its 'learning bias.’

How to combat AI bias?

• Broaden the Learning Material: we need to ensure the data AI learns from is varied and representative of real-world diversity.

• Constantly Review Learning: Regularly examine what AI is learning, akin to correct any misconceptions early on.

• Ensure Diversity in Development: Include a wide range of voices and perspectives in AI development teams.

THROUGH A-EYES EARS

Well, this was a revelation! Suno.ai enables you to make music, write lyrics, create songs, and generally have fun. It’s so easy too. Just sign up, you get 50 free credits - daily - and use a simple text prompt. Below are a couple of samples.

1 and 2 are random songs but I decided to make #3 a lighthearted tribute to AI 🤣 🙄 I prompted with the general idea and genre, added a few specifics - et voila!

FREE RESOURCES

CURIOUS CAT

Meet Jackson Barn Cat a.k.a. JBC, office manager, resident AI-whisperer, and all-around curious cat. Each week he’ll nose bop a new challenge 😺

Do you have a story about AI@Work that you’d like to share - maybe a new use case, some tips, or an unexpected AI success story? If so, we’d love to hear from you! [email protected]

If you know anyone who might enjoy more content like this, we’d be honoured if you would share this newsletter with them 🙂 Word of mouth is our best growth engine 💯 http://www.ai-foreveryone/subscribe