AI created more of a BUZZ in Cancun then the tequila. Ok...not true...but almost....

A DOSE OF WISDOM

"Do not wait to strike till the iron is hot, but make it hot by striking." – William Butler Yeats

THE FIRST HIT

Without a doubt, the biggest session and the highest level of interest at the recent Lenders One summit was all about AI.

It's abundantly clear that AI will have an impact on mortgage lending and in many different areas, the questions are, what are those areas, how can they be impacted, what are the economic benefits to doing so, what are the opportunity costs of not deploying AI in various areas of your companies, and.... what are the risks.

There is one aspect of the discussion surrounding AI that is not up for discussion and that is that if you are not ALREADY exploring how AI can impact your business, you are LATE to the party.

There are some things that AI is already currently very very good at, and it's use can have an instant impact on your business. AI OCR technologies, using AI to examine bank statements for font variations, ink variations, inconsistencies, or anything that may indicate fraud. As great as AI is at helping solve problems, it will create others....and the fraudsters have already figured out how powerful a tool AI is.....and they will use it to facilitate fraud. There are a number of areas where AI can improve processes, and increase efficiencies that are workflow related and that have real and immediate positive economic impacts on your business.

There was also a lot of discussion around Chatbots, and anything consumer facing as the risks associated with one of these Chatbots is REAL. The regulator formerly known as the CFPB has made it clear that any ChatBot and the information it disseminates will be viewed as having been disseminated by a REAL employee. In other words, if the ChatBot makes a mistake, you may have just bought yourself a lawsuit. Not only that, but in what appears to be a changing regulatory environment, it's looking more and more like a lot of that regulation is going to be left to the 50 states, as opposed to a federal agency. So, what do Texas, California, Florida, and New York have to say about how these things, and other aspects of AI are regulated???? How do you steer clear of any regulatory malfunctions?

Perhaps it makes more sense to employ AI in areas of your process that are relatively free from forward facing or decisioning...anything that has fair lending regulations, and focus on areas where perhaps the AI is a little more mature, and less likely to have any cost savings or revenue generation wiped out by a regulatory smackdown...at least until some of this regulatory uncertainty shakes out.

Airline held liable for its chatbot giving passenger bad advice - what this means for travelers - When Air Canada's chatbot gave incorrect information to a traveller, the airline argued its chatbot is "responsible for its own actions".

One of the biggest revelations for me in these sessions came form of the realization that there is actually "Explainable AI" that can be used in the decisioning process.

"Explainable AI" was a completely new term for me, and frankly sounded like an oxymoron. CLEARLY any decisioning done by AI HAS to be able to be justified and has to be done in a way that is free from discrimination or any fair lending malfunctions. FNMA and FHLMC have ALWAYS been pressured to show how their "black box" produced loan decisions. In traditional AI models, not only can they hallucinate, or introduce bias, the can also do so without transparency. That is NO BUENO!...

In the interest of transparency, in an effort to learn MORE about EXPLAINABLE AI, I queried ChatGPT. Here is that result. It's quite interesting.....

What is Explainable AI (XAI)?

Explainable AI (XAI) refers to artificial intelligence systems designed to provide clear, understandable, and interpretable outputs. Unlike traditional "black-box" AI models that make decisions without transparency, XAI focuses on ensuring that users can understand, trust, and interpret the reasoning behind an AI’s predictions or decisions.

Why is Explainable AI Important?

1. Trust & Transparency – Users need to trust AI decisions, especially in high-stakes industries like finance, healthcare, and law.
2. Bias Detection & Fairness – Helps identify and correct biases in AI models.
3. Regulatory Compliance – Laws like GDPR and the AI Act require AI systems to explain their decisions.
4. Better Decision-Making – Humans and AI can collaborate more effectively when AI outputs are understandable.
5. Debugging & Improvement – Engineers and data scientists can refine models when they know why a model makes a particular decision.

How Does Explainable AI Work?

XAI techniques help break down AI decisions into understandable components. Some common approaches include:

1. Feature Importance – Shows which variables (features) were most influential in the model’s decision.
2. SHAP (SHapley Additive Explanations) – Uses game theory to fairly assign credit to each input feature in a prediction.
3. LIME (Local Interpretable Model-agnostic Explanations) – Generates interpretable approximations of AI predictions for individual cases.
4. Decision Trees & Rule-Based Models – These models are naturally interpretable, making it easy to see the decision path.
5. Counterfactual Explanations – Answers the question, “What would need to change for a different outcome?”

Where is Explainable AI Used?

• Finance – AI-driven credit scoring and fraud detection require transparency.
• Healthcare – AI assisting in diagnoses needs explainability for doctors and patients.
• Legal & HR – AI-driven hiring and legal decisions must be fair and unbiased.
• Autonomous Vehicles – Understanding why an AI system chooses a particular driving action is critical.
• Customer Service – Chatbots and recommendation engines must explain their suggestions.

Challenges of XAI

• Trade-off Between Accuracy & Explainability – More complex models (like deep learning) are harder to explain.
• Security & Privacy Risks – Too much transparency can expose sensitive data or allow bad actors to manipulate AI systems.
• Human Interpretation – Even with XAI, non-technical users may struggle to understand complex explanations.

Final Thoughts

Explainable AI is essential for building AI systems that people trust and understand. As AI becomes more integrated into daily life, XAI will shape the future of ethical, fair, and accountable AI systems.

Homebuyers are making bigger down payments. Coming up with that cash is still a struggle

• In December, the median down payment among homebuyers was $63,188, up 7.5% from a year ago, according to a recent report by Redfin.
• But for most, coming up with a down payment is still a “very significant” hurdle, according to a new report by Bankrate.
• ANOTHER GREAT REASON TO USE CLICK N' CLOSE'S NATIONWIDE DPA with NO INCOME RESTRICTIONS, NO FIRST-TIME HOMEBUYER RESTRICTIONS, AVAILABLE AS A 3.5% OR 5% DPA AND HAS A SHARED APPRECIATION OPTION!!!!

Preventing Mortgage Fraud – Take a Closer Look

75 minutesIn this webinar, you'll learn about:

• What fraud is and its effects
• Where it frequently occurs
• Common schemes and motivations
• How red flags help assist in fraud detection
• How to ask the right questions to maximize your effectiveness and value to both your borrower and your company

The Federal Housing Administration could face layoffs — what that may mean for homebuyers - CNBC

Here’s how tariffs will hit the U.S. housing market

• Lumber, gypsum for drywall and home appliances are expected to be impacted by tariffs on Canada, Mexico and China.
• The new tariffs could increase builder costs anywhere from $7,500 to $10,000 per home, said Rob Dietz, chief economist at the National Association of Home Builders, citing estimates from U.S. homebuilders.
• The greatest impact to homebuilders will be from lumber cost increases, which are expected to total about $4,900 per home on average, according to Leading Builders of America.

Click n' Close SmartBuy Program - If you have not added this arrow to your quiver, you are not fully armed to attack the market. No income restrictions - No First-time Homebuyer restrictions, available NATIONWIDE (sorry NY), forgivable, non-forgivable and shared appreciate mortgage (SAM) options... and you can buy down the rate!!! Instead of adding a slew of different state DPA's add ONE!!!! One that can expand your customer base, help your customers tackle affordability issues AND increase your bottom line!!!

CONTROLLED MARKET SUBSTANCE

Trump Volatility - Yesterday the S&P 500 opened up, after a pretty brutal day the day before. It was chugging along, up about 20 points, when Trump comes out and says he is going to give the automakers a 30 day exemption from the tariffs and the stock market rallies another 50 points in 30 MINUTES.

Now.....it's being reported that he is considering agricultural exemptions for both Mexico and Canada. Meanwhile, China is basically saying "bring it on". What this amounts to, and what much of the general public doesn't seem to understand, is that much of this is essentially economic saber rattling. And.....unfortunately for Mexico and Canada, it's more like pen knife rattling. They, as it has been said recently....."have no cards'........

10 Year Yield - You know what tomorrow is!!! It's non-farm payroll Friday...AND JPOW speaks tomorrow afternoon!!! Betting rates won't be where they are today come Friday afternoon!