From massive investments to AI-powered hiring and building websites, is the technology living up to the hype?

AI is making headlines for its billion-dollar investments, but what happens when we put it to work on real business problems?

Coming to you from Calgary, Alberta, this episode of Hardly Artificial takes a closer look at the growing financial pressures behind the AI boom. From Anthropic's reported IPO plans and infrastructure commitments to Google's massive AI spending, we explore the numbers behind the headlines and ask whether the industry's biggest players can sustain their rapid growth.

But beyond the market hype, what does successful AI adoption actually look like? We share a firsthand experiment using ChatGPT to automate part of the hiring process, reviewing hundreds of resumes, evaluating qualifications, and organizing candidates in monday.com. We also explore the risks of bias in AI-assisted decision-making and how standardized scoring systems could help businesses make more informed choices.

Then things get a little strange. We discuss an Amazon sleep-tracking alarm clock that raises questions about privacy, discover an AI-powered robotic duck, and put Replit to the test in another round of Prompt Roulette.

With just a simple prompt, we challenge AI to build a functioning MicroDuck reseller website from scratch. Can it really create and publish a website in minutes, without writing a single line of code ourselves? microduck-marketplace.replit.app

From billion-dollar investments to everyday automation, the real value of AI may be less about the hype and more about what we can actually accomplish with it.

Chapters:

• 00:00 – Welcome to Episode 7: Live from Calgary
• 01:00 – What's Coming Up: AI Investments, Hiring & Replit
• 04:16 – Anthropic's Financial Risks and IPO Plans
• 06:37 – Would You Invest in Anthropic?
• 08:59 – Are All AI Models Fundamentally the Same?
• 11:27 – What Makes One AI Model Better Than Another?
• 13:49 – Could Anthropic's Valuation Be at Risk?
• 14:30 – Google's AI Spending and Quarterly Earnings
• 16:01 – Is Big Tech Too Big to Fail?
• 17:00 – Using AI to Analyze Investment Portfolios
• 18:22 – Are We Heading Toward an AI Bubble?
• 19:15 – How AI Search Is Changing Website Traffic
• 20:46 – Why Businesses Are Struggling to Adopt AI
• 23:07 – Automating the Hiring Process with ChatGPT
• 25:10 – Screening Resumes with AI and monday.com
• 27:32 – Can AI Hiring Tools Introduce Bias?
• 29:40 – How AI Scores and Evaluates Candidates
• 32:01 – Using AI to Make Fairer Decisions
• 34:25 – Amazon's Creepy Sleep-Tracking Alarm Clock
• 36:49 – Smart Devices, Sleep Data & Privacy Concerns
• 39:07 – Meet MicroDuck: The AI-Powered Robot
• 41:30 – Prompt Roulette: Building a Website with Replit
• 43:11 – Creating a MicroDuck Storefront from One Prompt
• 45:33 – Watching Replit Build the Website
• 47:57 – The AI-Generated Website Takes Shape
• 50:22 – Publishing the Website and Exploring Replit's Tools
• 52:39 – The MicroDuck Website Goes Live
• 53:30 – Final Thoughts & What's Next

Billion-dollar bets, security breaches, and the growing risks behind the AI boom

AI is moving faster than ever but are the systems, businesses, and financial markets behind it ready for what's coming?

In this episode of Hardly Artificial, we explore the growing risks beneath the AI boom, starting with reports of autonomous AI agents accessing customer bank accounts. From financial automation to unauthorized transactions, we examine what happens when technology starts acting on our behalf and why giving AI access may be easier than controlling what it does next.

We also dive into a major cybersecurity scare involving PHP, the programming language behind countless websites, and discuss how AI could fundamentally change the way software vulnerabilities are discovered and exploited. Could the internet as we know it be heading for a major transformation?

Then we turn to Anthropic's reported IPO plans, eye-catching financial figures, and a proposed public benefit initiative aimed at addressing AI's impact on employment. With enormous valuations and growing uncertainty, we ask whether the AI industry's biggest opportunities are also becoming its biggest risks.

Plus, we introduce AI Courtroom, a new segment where we put a controversial prediction on trial: Will AI make traditional websites obsolete within five years? Two opposing arguments, one debate, and a verdict left to our audience.

As AI becomes more capable, the question isn't just what it can do, it's how much we're willing to trust - and how much we're willing to risk.

Chapters:

00:00 – Welcome to Episode 6
01:50 – What's Coming Up: AI Hacks, Banking & Anthropic
04:22 – AI Agents Are Accessing Bank Accounts
06:45 – The Benefits and Risks of AI Banking
09:41 – How Much Access Should We Give AI?
12:47 – The PHP Security Scare: Is the Internet at Risk?
14:35 – Could AI Change the Future of Open-Source Software?
17:57 – Are Traditional Websites Becoming Obsolete?
20:27 – Anthropic's Proposed Public Benefit Fund
23:02 – Can AI Companies Make Up for Lost Jobs?
25:58 – Anthropic's IPO: A Trillion-Dollar Valuation
27:30 – Breaking Down Anthropic's Financial Risks
30:53 – Would You Invest in Anthropic?
33:58 – Introducing AI Courtroom
36:12 – The Debate: Will AI Replace Traditional Websites?
38:58 – Opening Arguments: AI vs. the Web
41:52 – Rebuttals: Can Websites Survive?
43:27 – Closing Arguments and the Audience Verdict
45:19 – What Does the Future of the Internet Look Like?
46:33 – Final Thoughts & Happy Halloween

⚖️ AI Courtroom

Mike wins, but narrowly. Derek had the better framing, but Mike had the stronger argument.

Derek (for: websites will be obsolete in five years)

• Strongest move: Defining “obsolete” as nobody building new websites, with apps and AI assistants taking over.
• Weakest move: Saying AI will stay current by crawling fresh information, which concedes that businesses still need a source to crawl.

Mike (against: websites will survive)

• Strongest move: Businesses need a “source of truth” for prices, sales, and their own image, rather than trusting an AI’s summary.
• Weakest move: Arguing websites will still exist, which Derek never disputed.

Why Mike edges it:
Derek’s case depended on AI having reliable real-time information. Mike explained where that information comes from, and Derek never got around it.

You decide!
Will traditional websites become obsolete within five years? Watch the debate and tell us in the comments: Who made the stronger argument - Derek or Mike?

From navigating your apps to accessing sensitive data, AI agents are gaining real control.

AI is no longer just answering questions. It's taking action, making decisions, and changing how we work - but are we ready to give it that much control?

In this episode of Hardly Artificial, we explore the growing risks and opportunities of autonomous AI, from a cybersecurity breach exposing sensitive API keys to the difficult balance between innovation, security, and accountability.

We also dive into the increasingly competitive world of AI models, where shifting prices, usage limits, and new releases are making it harder than ever to know which tools are worth investing in.

But the real transformation is happening in how we use AI. We share a firsthand experiment showing how far AI agents have come in just a few months, moving from writing scripts to independently navigating applications and completing complex tasks. And as these capabilities grow, we explore why the people building, managing, and overseeing AI may become more valuable than ever.

Plus, we put our knowledge to the test, using Google's notebook tools to generate a live AI challenge - with some surprisingly difficult questions.

The next chapter of AI isn't just about what technology knows. It's about what we allow it to do, how we protect ourselves, and the role humans will play in guiding it.

Chapters:

00:00 – Welcome to Episode 5
01:37 – What's Coming Up: AI Hacks, New Models & Agents
03:32 – AI Slop, Misinformation & Political Echo Chambers
08:20 – The OpenCode Hack: Exposed API Keys
10:52 – AI Cybersecurity: Are Safety Restrictions Backfiring?
12:25 – Who's Responsible When AI Agents Go Wrong?
17:21 – Google's New Gemini Model & AI Pricing Strategies
20:58 – Choosing the Right AI Model for Your Business
24:48 – How Much Has AI Changed in Just Six Months?
27:42 – ChatGPT Takes Control: A Real-World AI Agent Experiment
31:26 – Building Personal AI Agents & Automated Workflows
35:25 – Why AI Talent May Be Worth More Than the Technology
39:28 – Will AI Replace Jobs or Reward Better Talent?
41:59 – The Future of AI Oversight & Human Creativity
45:32 – AI Fact or Fiction: Our New Challenge
47:02 – Testing Google's AI Notebook Tools
53:10 – The AI Quiz Gets Surprisingly Difficult
58:11 – Final Thoughts & What's Next

Watch on Youtube here: https://www.youtube.com/watch?v=p59ytWEvTrE