- AI&TFOW
- Posts
- From Extraction to Understanding: Next Era of Document AI [Newsletter #51]
From Extraction to Understanding: Next Era of Document AI [Newsletter #51]
Why AGI Is The Wrong Goal
AI and the Future of Work
June 03, 2025
Hello, AI enthusiasts from all around the world! Welcome to the weekly newsletter for the AI and The Future of Work Podcast.
The world is rushing to create a replacement for human intelligence when that should not be the goal.
Today's guest urges us to rethink our quest around artificial intelligence. After all, humans should be the focus, regardless of the product we're creating.
That's why sometimes it's best to focus on the boring side of AI, to uncover our true potential.
Let's dive into this week's highlights! 🚀
🎙️ New Podcast Episode With Martin Goodson, CEO of Evolution AI
Why are we trying to recreate human intelligence as if there were a shortage of it? This quest has always intrigued Martin Goodson, but not because he wants to take part. Instead, he sees no reason for it.
"There is not a deficiency of intelligence," he highlights. There are billions of intelligent humans out there. So we must ask ourselves whether this is the right goal. Should we race to replicate something that already exists and (most importantly) is abundant?
The real challenge with today's society isn't about replacing intelligence, although that idea has become popular. Instead, Goodson wants to focus on the boring side of AI.
We've all read the sci-fi novels that envision extremely advanced creations and it's okay to dream. No one wants to read 4 million pages of bank statements, but machines can.
What’s more important: they should.
That's the AI we need—something that complements our roles instead of replacing them.
That’s been Goodson's goal since he launched Evolution AI in 2012: to apply deep learning to optical character recognition (OCR).
Interestingly, it all started with a task many might find boring: taxes. Before founding Evolution AI, he made several attempts to use OCR to read tax documents and calculate how much people owed.
He ran into an interesting surprise. OCR did its job: it read pages well but could not process information efficiently.
For that, he needed something else. So he made it his goal to merge OCR with deep learning so that machines could help humans understand and process documents faster.
As a result, Evolution AI has received one of the most significant AI R&D grants ever awarded by the UK government, along with investment from First Minute Capital.
A former scientific researcher at Oxford University, Goodson has led AI research across several organizations and was elected Chair of the Data Science and AI Section of the Royal Statistical Society in 2019.
He sat down with Dan Turchin, PeopleReign CEO, to explain how OCR missed the mark and how combining visual and linguistic context unlocked a new level of document intelligence, plus much more:
Why we should build systems that support human thinking, and not just mimic it.
What Martin believes intelligence really is, and why language alone may be the wrong benchmark for AGI.
The evolution from data extraction to accurate financial analysis, and why domain knowledge is essential for reading statements like income reports.
The risks of LLM hallucinations, especially with numerical data, and why accuracy still requires combining techniques across model types.
🎧Listen to the full episode here to learn more about Goodson's vision for the future of Artificial General Intelligence, and why he believes in the value of defining intelligence in the simplest of terms: the ability to think and learn.
This fascinating conversation also mentioned two previous episodes.
Here, you can listen to our conversation with Daniel Marcous on doing taxes in the era of AI.
Then, you can listen to Peter Voss here on making AI smarter without harming humans.
📖 AI Fun Fact Article
Tiger Brokers is an online broker with backers like Chinese electronics manufacturer Xiaomi and US investor Jim Rogers, big names in their respective industries.
So when this big-name broker embraces DeepSeek, China's ChatGPT competitor, the world takes notice, and these aren’t isolated incidents.
At least 20 financial firms, including Sinolink Securities, CICC Wealth Management, and China Universal Asset Management, are integrating DeepSeek models into their businesses.
As Samuel Shen and Tom Westbrook write in Reuters Online, these financial institutions are accelerating their use of AI in research, risk management, and investment strategies. The shift reflects China's broader push for AI-driven financial innovation, a trend the authors call out as carrying geopolitical implications.
Source: Reuters
China's rapid deployment of AI in finance challenges the current Western dominance in global Fintech and data-driven capital markets.
Dan Turchin, PeopleReign CEO, urges caution on this statement. At this time, the geopolitics of AI are tugging at sensitive topics, from chip restrictions to how accessible (aka, free vs. paid) these platforms will be.
These topics will need solving soon. After all, foundation models want to be free.
It's unrealistic to think the best models won't soon be freely available and accessible to all countries, regardless of how any administration uses tactics like trade barriers and economic sanctions to deter access.
Let's prepare for and celebrate a world where the most ambitious technologists and entrepreneurs can solve the most complex problems unimpeded. Responsible and fair use of AI will be a key part of making that a reality.
Upcoming LinkedIn Live: The Evolving Role of People Leaders
📅 Tuesday, June 10 | 9:00 AM PDT | Join
Don’t miss Dan Turchin, CEO of PeopleReign and host of AI and the Future of Work, live in conversation with TalentLign CEO Pramukh Jeyathilak.
This session will explore how modern people leaders are reimagining their roles by blending human-centered leadership with data and technology-driven innovation.
Dan has spent over two decades building AI-first platforms for HR and IT at companies like Astound, InsightFinder, and ServiceNow. He’s a startup founder, product strategist, and investor known for using AI to solve real-world problems.
Dan will bring his deep AI expertise to the table, sharing how technology is helping HR leaders make smarter decisions without losing the human touch.
Topics include:
Trust, belonging, and culture in a tech-driven world
Why employee experience is no longer just HR’s job
How AI is helping leaders move with clarity and purpose
Listener Spotlight
May, from Nashville, Tennessee, is in customer support for a tech company and listens while on her break.
May's favorite episode is the great one with Daphne Jones, author of Win When They Say You Won't, back in season 3, about overcoming imposter syndrome.
You can listen to this excellent conversation here!
🎥 Worth a Watch
We've all heard that the AI revolution is overhyped. Many even believe we're reaching a bubble.
So, why is Google's former CEO, Eric Schmidt, saying that the AI revolution is actually underhyped?
The staggering number of opportunities alone is worthy of substantial hype, as we're most likely underestimating how much we can do with AI.
Check out this fascinating conversation where he addresses this while also reminding us of sober realities around AI.
We want to hear what you have to say! Your feedback helps us improve and ensures we continue to deliver valuable insights to our podcast listeners. 👇
Until next time, stay curious! 🤔
We want you to stay informed about the latest happenings in AI, so we curate important news worldwide.
This country is regarded as the European educational powerhouse and has now integrated artificial intelligence into its education program.
This giant partnered with OpenAI, NVidia, and xAI, which has many wondering if it has too much power.
Why is making AI code-production ready so difficult? This article explains.
Until next time, stay curious! 🤔
Martin Goodson likes to use AI to tackle the tasks that many of us might find boring. Making those tasks easier can help thousands have a more productive, healthier work-life balance.
It's one of the most impactful uses of AI, and it certainly opened our eyes to rethinking problem-solving in the future.
We hope it has done the same for you. Until next time, keep questioning, keep innovating, and we'll see you in the future of work! 🎙️✨
If you liked this newsletter, share it with your friends!
If this email was forwarded to you, subscribe to our Linkedin’s newsletter here to get the newsletter every week.