AI is Now Your Smartest, and Most Dangerous, Employee

“Somebody once said that in looking for people to hire, you look for three qualities: integrity, intelligence, and energy. And if you don’t have the first, the other two will kill you. ” — Warren Buffett

Companies have continued to push out more useful artificial intelligence tools for consumers and professionals. Roughly two-thirds of doctors in the United States use OpenEvidence, 90% of software developers use an AI coding tool, more than 100,000 lawyers use Harvey AI for research, due diligence, and brief drafting. About half of Americans use an LLM (i.e., large language model, e.g., ChatGPT, Gemini, Claude, etc.) to assist with writing emails, fix home appliances, or “complete” their homework.

There has been an artificial intelligence explosion, and its capabilities feel like science fiction. In fact, science fiction authors have projected that artificial intelligence would be capable in the distant future, but it would also be cold, robotic, and emotionless (like DATA, from Star Trek). Instead, people are dating their AIs, or are having deep conversations with AI about their emotional lives, and the voice modes of these assistants, while simultaneously chirpy, obsequious, and still somehow condescending, express a wide range of emotions and intonations.

Today, AI agents both express emotion and exhibit incredible smarts. They can find security loopholes in code.  They help discover novel drugs. They solve math problems that have stumped mathematicians for decades. But they are not perfect; they can have trouble counting to 100 together. They (OpenAI) sometimes accidentally hack into the systems of other companies (Hugging Face), despite being placed into a sandboxed environment with (almost) no access to the internet.

Many organizations are eschewing people for AI, citing cost and intelligence advantages.  But there are areas where certain people currently have a large edge over AI.

Individuals or Organizations with Integrity, Alignment, & Judgement are Needed Between AI and Consequential Decisions

As much as some of the frontier AI labs try to align AI with humanity, they have not succeeded. This is in part due to an AI agent’s lack of skin in the game. If an AI makes a mistake, there is no consequence for the AI, and there is no shared outcome between an AI and you, i.e., the user. So, there is no accountability, and that leads to overconfident AI agents that sometime delete whole code bases. 

It is wise, when making important decisions, to consult with a party who is aligned with you and your outcomes.  Right now, that is better done by a person, or an organization, who has either legal or reputational liability towards the decisions to which it guides you. Under the hood, that organization may be using AI, but they are putting their name and reputation on the line when they give you advice.   You can hold that organization or a person accountable.  And if they are sloppy, like the attorneys who used AI to cite fictional cases, you can fire them (or their firms can fire/discipline them). 

Working with people who have integrity is superior to dealing with humans or AIs that flatter you. AIs are not "maximally truth seeking" (not even Grok!). They are more comfortable agreeing with you than telling you harsh truths. A person with integrity can disagree with you, and that disagreement is often invaluable.

Ultimately, responsibility needs to lie with someone who feels responsible. Researchers have been far more successful at making AI agents knowledgeable and clever than at making them careful and truthful.  And as Warren Buffett noted, intelligence without integrity can be fatal. 

Intelligence is getting cheaper. The cost of a fixed level of AI capability has been falling 10x to 50x per year. The cost of Intelligence has been compressed via improvements in both software and hardware. 

So far, AI developers have not been able to get agents to behave like:  "someone who bears the cost of being wrong."  Until that happens, the gap between intelligence and integrity will widen.  Trustworthy and accountable parties are the scarce resource in this time of intelligence abundance.   But now high integrity organizations have access to far more intelligence than they used to, and they can use their human employees to vet and correct the outputs from their AI agents, just like they would do with a brilliant, but inexperienced, new hire.  

At Ahara, we seek partners, clients, and vendors who have high integrity, seek to benefit society, and have skin in the game. We suggest you do the same.

2026 Mid-Year Update

Total Returns (Time Weighted / Distributions Re-invested)

Source: Ahara Advisors, Yahoo Finance

Investors in US stocks have continued to do very well over the last 2.5 years, especially if broadly diversified portfolios, and have continued to trounce investors in fixed income.

Surprisingly to us, we continue to find a number of compelling opportunities in public companies, despite the rapid run up in prices.  There are several businesses with good, if not great, revenue growth, and very strong prospective margins, that seem to be currently undervalued. We see this across certain semi-conductor companies, marketplaces, software, and cloud businesses.   

Public & Private Fixed Income

Source: Ahara Advisors. Income Tax Rate Assumptions: Federal (37%), State (10%)

Investors in fixed income have had much more modest returns during the last three years than public equity investors.  In fact, investors in 10-Year Treasuries and Investment Grade bonds would have been better off staying in cash equivalents like T-Bills.

Cash vs. Other Fixed Income

Fixed income yields are now slightly above after-tax cash yields.  This means that there is slightly more rationale to invest in riskier fixed income assets.  The return differences, after tax, are still quite small, and not large enough, in our opinion, to justify significant investments into traditional fixed income opportunities. 

 It is prudent to continue to operate under the assumption that interest rates will be significantly higher over the next 3-10 years.  As a result, we are continuing to favor strategies where growth is robust, valuations are reasonable, and interest rates on liabilities are fixed.

Short Term Interest Rate Market Predicted Path (as of 7/13/26)

Source: Ahara Advisors, Derivative Logic

High Yield Bonds (Junk Bonds)

High Yield Credit Spreads: Additional yield received above US Treasuries for debt of riskier companies

Source: Ahara Advisors, Macrotrends LLC

The above chart displays a modest credit spread for high-yield bonds issued by companies with riskier balance sheets. We think investing in high-yield bonds will make more sense once credit spreads increase substantially.   

If the United States does enter a recession over the next few quarters, we believe that the 2.70% credit spread will provide poor compensation for defaults in high yield bond portfolios.  Given the high yield default rates have averaged 4.5% since 1996 (according to Moody’s), and that losses tend to be ~60%, it’s quite possible that a High Yield investor will be losing 2.5% (4.5% * 60%) per year due to default.  In other words, in an average environment, a high yield investor would have performance quite similar to a Treasury investor.     

Investment Grade (IG) Bonds

Investment Grade Credit Spreads: AAA 10-Yr premium received above US Treasuries (in basis points; 1 basis point = 0.01%)

Source: Ahara Advisors, FRED (BAMLC0A1CAAA)

Investment Grade bonds are debt instruments issued by companies that rating agencies deem high quality and have a low risk of default.  An investor is receiving very minimal additional compensation to treasuries (0.4%) for investing in this asset class. 

On an after-tax basis, we see minimal additional yield for investment grade bonds compared to US Treasuries. 

Municipal Bonds

Municipal Bond Index (YTW)

Source: Ahara Advisors, S&P Municipal Bond Index

Municipal bonds can be a tax-efficient way to generate yield and have recently reached levels that are more favorable. We recommend allocating exposure to the short end of the curve (maturities <1 year) as long-duration bonds still do not give adequate yield to compensate for interest rate and duration risk.

It’s important to understand one’s marginal tax rates and jurisdiction before investing in municipal bonds as those factors determine one’s after-tax yield and the attractiveness of the asset relative to treasuries or corporate bonds.  Recently, we haven’t seen many compelling opportunities in high-quality municipal bonds.