Evaluating the Value of Four-Transcript Analyses for Biz Deals

 Recently we looked at "four investor call" analyses for Bioreference/Labcorp, Invitae/Bankruptcy, and Myriad/Labcorp.  Quo vadis?

##

The value of “Four Transcripts” is that it lets us test a company’s explanation against its own subsequent results. One call offers a carefully prepared account of the present and a forecast of the future. Four calls let us ask whether the forecast arrived, whether the explanation changed, and whether the original business objective still made sense.

Across BioReference, Invitae, and Myriad Prenatal, that approach produced three different findings. Their common destination—Labcorp—should not obscure how differently they arrived there.

Three cases, three different problems

Case	Management’s central proposition	What the sequence revealed	Principal lesson
BioReference	A smaller, more focused laboratory could achieve sustainable profitability	Operating improvements were real, but recurring profitability remained unproven; a transaction-related benefit complicated the apparent earnings improvement	Improving a business does not establish that its current parent is its best owner
Invitae	Restructuring and refinancing could provide enough time to reach a sustainable business model	Margins improved and cash burn fell, yet financing needs and the capital structure remained decisive	An operating turnaround can progress while the corporation’s ability to finance it deteriorates
Myriad Prenatal	Customer recovery, dedicated selling, and FirstGene could restore growth	Recovery disappointed, explanations broadened, and prenatal increasingly competed with oncology for resources	A promising business can lose its place in a company’s investment priorities without becoming technologically obsolete

These are interpretations of the three supplied studies, rather than a new independent audit of every underlying transcript and filing.

BioReference: the revealing feature was persistence. Its public strategy remained remarkably consistent: reduce costs, concentrate the business, improve the testing mix, and grow 4Kscore. The useful observation was how long the anticipated profitability remained ahead of the company.

The report’s most concrete analytical finding was the distinction between reported and recurring operating performance. A reported quarterly operating profit included a divestiture-related benefit large enough to reverse the apparent conclusion about the continuing operation. That is something an enthusiastic call summary could easily miss.

The sale did not prove that the turnaround was fictitious. Restructuring could improve both the prospects for continued ownership and the price obtainable from a buyer. The calls demonstrated an operational improvement program; they did not disclose when selling became the preferred corporate strategy.

The BioReference study provides that distinction: BioReference_Terminal_Year_OPKO_Case_Study_Updated_PAMA(2).pdf

Invitae: the revealing feature was the coexistence of improvement and danger. This is the strongest answer to the claim that earnings calls are merely hype. Some of the positive operating developments were substantive. Gross margins improved, spending fell, and near-term debt maturities were addressed. Those achievements nevertheless failed to establish that the company could finance itself through to sustainability.

The decisive change was the transition from a turnaround supported by expected financing flexibility to a formal going-concern warning and a capital-structure review. The report carefully distinguishes that management assessment from a supposed longstanding auditor warning.

The important analytical question was therefore not whether management could identify genuine progress. It could. The question was whether that progress solved the company’s binding financial constraint.

The Invitae study captures that separation: INVITAE Lst Four Calls.pdf

Myriad Prenatal: the revealing feature was the evolution of the explanation. The early account centered on an ordering-system disruption and recovering displaced customer volume. Later calls placed increasing weight on dedicated sales resources, competition, reimbursement, and FirstGene. Finally, portfolio review became explicit.

Your challenge to the software explanation improved the analysis. Repetition by management is evidence that an explanation mattered to its public narrative. It is not an independent measurement of how much that factor contributed to the outcome.

The stronger finding was that the recovery thesis became less convincing while the opportunity cost of continuing to fund prenatal became more important. The sale could make sense even if FirstGene was scientifically sound and commercially promising.

The Myriad study makes that distinction explicit: 20261020 Myriad to Labcorp Prenatal.pdf

“The calls are just BS and hype”—a useful warning, but an inadequate method

Earnings calls are managed communications. Executives choose which metrics to emphasize, which comparisons to make, and how to describe setbacks. A favorable statement can be accurate while leaving the reader with an incomplete understanding.

That is precisely why reading a sequence can be productive.

Consider three hypothetical statements:

• “Our gross margin improved.”
• “Our restructuring is ahead of plan.”
• “We expect growth to resume in the second half.”

The first may describe a measurable achievement. The second requires a definition of the plan and its baseline. The third is a forecast that can be checked later. Treating all three as either trustworthy information or empty promotion discards the distinctions that make analysis possible.

The most informative question is often: What does this favorable fact actually establish?

Higher gross margin does not establish adequate cash generation. Lower cash burn does not establish sufficient financing. A successful product launch does not establish physician adoption, reimbursement, or attractive returns. An improving subsidiary does not establish that retaining it is the parent’s best use of capital.

The three cases become interesting when we stop asking whether management was “positive” and examine the distance between its evidence and its conclusion.

Why four calls add more than four separate summaries

A sequence creates several tests that a single transcript cannot provide.

First, it creates a record of commitments. A forecast made in one quarter becomes a question for the next. Did the milestone occur? Was its deadline changed? Was the metric redefined? Did management explain the difference?

Second, it reveals how the explanation develops. A temporary disruption may become a customer-retention problem, then a commercial-execution problem, then a portfolio decision. That progression might reflect better understanding, changing conditions, selective presentation, or some combination. It deserves investigation without an automatic accusation of dishonesty.

Third, it makes persistent language informative. BioReference illustrates this particularly well. There need not be a dramatic rhetorical reversal. Repeatedly approaching the same objective without clearly reaching it can itself become the finding.

Fourth, it reveals which questions become harder to answer. Analyst Q&A is valuable when someone asks for a reconciliation: Why has recovery not occurred? How much financing is actually accessible? Does improved reported profit represent the continuing business? Analysts are not infallible, but their questions can expose the assumptions that prepared remarks leave implicit.

Fifth, it helps distinguish a business’s performance from its owner’s choices. The relevant unit of analysis can shift from a test, to a laboratory, to a corporate portfolio, to the claims of creditors and shareholders. Invitae makes the danger of collapsing those levels especially clear.

The deepest commonality: progress was not enough

Each company could point to improvements. Each still faced a different unresolved question:

• BioReference: Would the retained operation generate durable earnings sufficient to justify continued ownership?
• Invitae: Could the company finance the remaining journey to sustainability?
• Myriad: Did prenatal offer a better use of incremental resources than the alternatives within the portfolio?

This suggests a useful discipline for any future four-call study: identify the constraint that matters most, then ask whether the reported progress actually relaxes it.

A company can spend considerable time discussing an improving metric that is relevant but insufficient. That need not be deception. Management may sincerely believe its achievements will eventually solve the larger problem. The analyst’s job is to examine the missing connection.

Labcorp gives the three cases a second analytical dimension

The buyer’s presence allows us to distinguish the value of operating assets from the success of the corporate structure that previously held them.

A laboratory can possess valuable assays, customer relationships, personnel, and data while struggling to generate satisfactory returns under a particular owner. A larger operator may anticipate using those assets differently or supporting them at lower incremental cost.

But the three outcomes should not be flattened into a single consolidation story. A negotiated divestiture and a bankruptcy sale occur under different constraints. The seller’s time, financing alternatives, and bargaining position matter. Purchase prices also cover different assets and liabilities.

Nor does Labcorp’s willingness to buy prove that its integration thesis will succeed. These cases document why ownership changed. A subsequent study would be needed to evaluate whether the new ownership delivered the expected economics.

The method needs safeguards against hindsight

The phrase “last four transcripts” is inherently retrospective. We selected these windows because we already knew something consequential happened afterward. That creates a temptation to make every earlier remark sound prophetic.

A fair study should ask two separate questions:

1. What could a careful reader reasonably have concluded at that time?
2. What does the later outcome help us understand about the earlier record?

The first supports contemporaneous judgment. The second supports historical explanation. They should not be presented as interchangeable.

Our three cases demonstrate explanatory usefulness; they do not establish predictive accuracy. To test prediction, we would also need companies with similar warnings or disappointing quarters that recovered, remained independent, or never sold. Otherwise, we cannot estimate false alarms.

Four calls are also a practical boundary, not a scientifically validated optimum. They provide repeated checkpoints, but acquisition history, financing terms, or reimbursement changes may require looking further back.

The strongest version is “four transcripts plus corroboration.” The calls organize the inquiry. Filings test liquidity and accounting claims. Transaction documents clarify what was sold. Product and payer evidence test commercial assumptions. Bankruptcy records can reveal constraints absent from earlier presentations.

That is also where AI is useful: aligning statements across quarters, finding changed forecasts, separating recurring from transaction-related figures, and organizing competing explanations. Its particular risk is producing an overly smooth story. The discipline is to preserve unresolved questions and label inference as inference.

The worthwhile lesson is that public corporate language can be analyzed without either believing it wholesale or dismissing it wholesale. These three studies found genuine progress, incomplete recoveries, shifting priorities, and materially different financial constraints. Reading the calls together made those distinctions visible—and made the eventual transactions more understandable without pretending they had been inevitable.

##

##

For Myriad you also hunted down an SEC document.

##

Yes—the October 20 Form 8-K and its attached asset purchase agreement, signed October 19. Those added information beyond the press release:

• Myriad retains pre-closing accounts receivable, so selling the business does not eliminate its responsibility for collecting those balances.
• The transaction includes employee offers, a patent license, and transition services.
• The agreement specifies closing conditions, potential extensions, and a $10.5 million termination fee in certain regulatory circumstances. www.sec.gov

That improves our description of the method: the four transcripts provide the narrative spine; the SEC documents test and deepen the interpretation. The calls show how management explained the business over time. The agreement shows what the parties actually committed to transfer, retain, and undertake.

We also found the historical SEC acquisition disclosure that distinguished Counsyl’s $375 million announcement price from $405.9 million in subsequently reported consideration—another useful detail that a press-release-only review would miss. www.sec.gov

Maverick Blue Line - East Boston

 

Rise and Fall of Wang Laboratories (Boston)

Was the decline of WANG comparable ot BLACKBERRY?

#

Yes! Wang Laboratories was one of the great Boston-area technology success stories of the 1970s and 1980s—and its collapse is remarkably reminiscent of BlackBerry's.

In fact, Wang is a classic example of a company that dominated a major technology market, became enormously successful, and then failed to adapt when the entire market changed underneath it.

The rise and fall of Wang Laboratories

7

Founded in 1951 by Chinese-born engineer An Wang, Wang Laboratories eventually became a Massachusetts technology giant headquartered in Lowell, about 25 miles northwest of Boston.

Its specialty was dedicated word-processing systems, particularly the Wang OIS (Office Information System). In the late 1970s and early 1980s, these were sophisticated, expensive office machines that could store, edit, and print documents long before ordinary office workers had personal computers.

Wang became especially popular with law firms, government agencies, corporations, and other organizations producing mountains of documents.

At its peak, Wang employed roughly 30,000 people and generated more than $3 billion in annual revenue.

How quickly did it collapse?

1984

Highly profitable: $210 million profit on $2.1 billion revenue.

1988

Revenue peaks around $3 billion, but the underlying business is deteriorating.

1989

Loses $424 million. An Wang removes his son Frederick as president.

1992

Files for Chapter 11 bankruptcy, with employment sharply reduced.

1999

The reorganized software and services business is acquired by Getronics.

Sources:

Los Angeles Times

+2

Why the BlackBerry analogy works

The similarity is almost uncanny.

Wang: Why would a corporation abandon an excellent $10,000–$20,000 dedicated word-processing system for a relatively primitive IBM PC?

BlackBerry: Why would a business executive abandon a superb physical keyboard and secure corporate email for an iPhone without a keyboard?

In both cases, the incumbent had a very good answer about why its existing product was superior—while missing the much larger advantages of the new platform.

There is one important complication: by the 1980s, Wang also had a substantial minicomputer business. Its failure was not simply the disappearance of dedicated word processors, but the erosion of its entire proprietary computing ecosystem.

And there is a fascinating Boston connection. Wang's decline was part of the collapse of the Route 128 computer industry, including Digital Equipment Corporation, Data General, and Prime Computer. Boston had once been a formidable competitor to Silicon Valley in computing hardware.

The Washington Post

+1

One surviving reminder is Boston's famous Wang Theatre, named for An Wang, who helped finance its restoration.

The historical irony: Wang helped invent the modern electronic office, but the personal computer revolution made Wang's own version of that office obsolete.

 

#

An Wang himself did not simply disappear from the scene. He remained involved with Wang Laboratories almost until his death.

He died March 24, 1990, at age 70, of esophageal cancer at Massachusetts General Hospital.

The Washington Post

+1

His story has a rather poignant ending:

• 1984: His company was still enormously profitable.

• 1986: His son Frederick became president, while An Wang remained CEO.

• 1989: With the company losing hundreds of millions, An Wang removed his own son from the presidency.

• 1990: Wang died, still chairman and CEO.

• 1992: The company filed for bankruptcy.

So Wang lived long enough to witness the unraveling of his life's work, but not its final bankruptcy.

He was also a remarkable inventor, earning a Harvard PhD in applied physics and making an important early contribution to magnetic-core computer memory.

And his philanthropic legacy outlived the company, particularly Boston's Wang Theatre and the Wang Ambulatory Care Center at MGH.

HOLLIS for Archival Discovery

+1

The striking historical detail is that the company went bankrupt just 29 months after its founder died.

Mark Twain as Movie Critic: Sound of Music, Love Actually

 

The Sound of Music

I had supposed that a convent existed to protect its inmates from the world. In this picture it serves chiefly to protect the world from a young woman named Maria, and presently abandons even that duty.

She is dispatched to a household containing seven children and a retired naval captain, an arrangement which would seem to require either a governess or an armistice.

The captain governs by whistle. Maria governs by song. The children, having examined both constitutions, elect the one which permits excursions.

Miss Andrews sings beautifully. Indeed, she sings so beautifully that one almost forgives the frequency with which she discovers an occasion to do so.

The baroness is the only person present who appears equipped to conduct an ordinary afternoon. Naturally, she must go.

Meanwhile Austria is being annexed, a development which introduces some welcome uncertainty into a plot whose matrimonial outcome has been visible from Switzerland.

At last the family escapes over the mountains. The youngest children manage this remarkably well, considering that their principal training has been in staircases.

It is a handsome and affecting entertainment. I left with renewed faith in children, mountains, and the wisdom of keeping a governess’s references on file.

Love Actually

This picture undertakes to prove that love is everywhere, and employs nearly the entire British acting profession to prevent the proposition from escaping examination.

There are so many courtships that I began assigning them numbers, a system defeated when several persons took an interest in the wrong number.

The prime minister falls in love with a member of his staff. In America this would require a committee; in England it requires a Christmas concert.

A gentleman courts his best friend’s wife by exhibiting written statements at her doorstep. He has evidently mistaken the absence of spoken testimony for innocence.

Another gentleman proposes marriage in Portuguese, a language he has recently acquired for the purpose. I admired his economy. Most husbands spend years learning how to be misunderstood.

Miss Thompson supplies the picture with its conscience. Her husband supplies it with the need for one.

The airport scenes suggest that the authorities will waive any regulation upon satisfactory evidence of a romantic emergency.

I laughed frequently and was moved against my better preparations. The machinery is shameless, but well tended.

I cannot certify that love is everywhere. In this picture, however, it has secured all the available parking.

Hemingway as Film Critic; Sound of Music, Love Actually

  

The Sound of Music

The girl came over the hill singing. It was a good hill and had done nothing to deserve this.

Below was the house of a captain. He had seven children and a whistle. He had been in the navy of a country which had misplaced its sea. A man must expect losses.

The girl made clothes from the curtains. After that there was no keeping the children indoors.

Julie Andrews has a clear voice and considerable endurance. Christopher Plummer looks like a man who has ordered a drink three times and been brought another song.  

There is a baroness. She understands money and chairs and sitting quietly. The picture treats her badly.

At last the Nazis arrive. They are against the family, but this does not make them the villains. They have other qualifications.

The family escapes across the mountains. By then you wish them well. You would have helped carry the children if it would get them out of this movie.

The photography is excellent. The mountains were good. They did not sing back.

 

Love Actually

In London at Christmas. There are too many people to follow, so the director gives some of them hats.

The prime minister loves the woman who brings him tea. He dances through government buildings. There has been no dancing at the embassy in Madrid. Things have gone badly enough without it.  The bulls.

A writer goes to France and falls for a Portuguese woman. They cannot understand each other.  

Alan Rickman buys a necklace for a woman who is not his wife. His wife receives a record. Emma Thompson makes the difference clear without explaining it. It is fine acting and belongs in a harder film.

One man declares his love with cards, avoiding conversation. Unfortunately, he has chosen his friend’s wife, so conversation will eventually be necessary.

There is a boy and an airport. Airport security is defeated.  (Must tell Orwell.)

Bill Nighy understands the film. He wants to sell a record and have a drink.

I recommend it. There is good work here, provided you wait for certain couples to leave.

Law Firms and AI

 dorf on law  https://abovethelaw.com/2026/10/biglaw-firm-will-pay-associates-to-be-bad-at-ai-for-a-while-100-hours-to-be-precise/   

tldr at bottom 

Biglaw Firm Will Pay Associates To Be Bad At AI For A While (100 Hours, To Be Precise)

It's the only way to learn AI, and this firm has figured that out.

By Joe Patrice on October 09, 2026 12:14 pm

There’s a chasm between understanding how to ask the free version of ChatGPT to punch up your emails and building a useful AI workflow. A couple weeks ago, I went to a conference with a panel titled “The Attorney’s Guide to AI” that was supposed to open lawyers’ eyes to building skills and deploying agents, but had to pivot to “so, you type in ‘c-l-a-u-d-e-DOT-a-i’ and enroll in a paid account.” For a change, this isn’t just a case of lawyers being hopelessly behind the curve on technology. The AI industry that needs to produce $6 trillion in revenue by 2030 to stay afloat only musters about $200 billion right now, mostly because the 98 percent of the public doesn’t see anything about AI to justify paying more than… nothing.

But for lawyers who understand how to elevate AI from the free sample platter to something that can provide actual value, they need time to experiment. The message from that same panel mentioned above was that every lawyer needs to start playing around, building skills, running tests, and figuring out what produces quality output.

Ropes & Gray understands this, and will now let all of its associates count up to 100 hours of AI experimentation toward their billable targets. It expands last year’s pilot, which allowed first-years to put up to 400 of their 1,900 hours toward playing with the robots.

This is the right call whether you expect generative AI to become the future of the profession or an ultimately minor productivity lift. Whatever the merits of the technology, moving into the world of skills, agents, and vibecoding is disorienting and the only cure is setting aside time to experiment with it. Whatever the plan, it’s going to fail in some spectacularly unexpected way on the first attempt. But if users keep iterating with the tool, it really does get better.

Just like a first-year associate.

One Ropes associate told Reuters, “you’re investing a lot of time in it and you’re not seeing a ton of efficiencies at first.” That might be an understatement. Horror stories of hallucinations and sloppy drafting tend to arise from practitioners trying to use cheap AI as an “Easy” button. Vendors building professional legal tools have done a lot of work to tailor it to a lawyer’s workflow. And it really is about “workflow.” Even if the technology provides accurate results out of the box, that doesn’t matter much if the way it does this robs the lawyer of meaningful judgment. The human brain can’t move faster and if the AI process isn’t giving the lawyer the necessary time to get the optimal as opposed to bare minimum result, that’s a workflow problem better addressed before it becomes the client’s problem. With individual tweaking, it can get even better.

As long as attorneys don’t become stereotypical “screamer” partners.

This policy generated some snide criticism, with many tying the move to recent reports that the big AI providers — especially Anthropic — have leaned on religious leaders to accept their belief that AI may be a conscious life form. Which is of course goofy.

That said, people who thrill in abusing AI have more than a little of that “serial killers always start out abusing animals” energy. Along with terms like “clanker,” which seem to flirt with people seeking the dopamine hit of using racial slurs without actually using any, there may be reason to crack down on treating AI as a sandbox for humanity’s worst impulses beyond whether it’s sentient or a fancy word calculator. And for the record, it’s the latter.

Look at me going off on tangents again.

Unfortunately, the billable hour treats experimentation as a personal failing. Any associate who wants to commit a Tuesday afternoon to building skills knows they’re losing time. Like sands through the hourglass so are our bonus targets. Giving associates time to learn AI without sacrificing their compensation will pay dividends down the road. It’s a mechanism that Ropes has deployed in the past, like when it let associates bill up to 100 hours for diversity and inclusion work. Of course, that was in 2021… before they deleted all the DEI info from their website to avoid offending the Trump administration.

Akerman has a more formalized process for giving credit for AI experimentation. At Akerman, playing with AI is another assignment that associates can pick up.

Either way, these firms are making an investment. While the AI cult wants you to believe the technology will soon replace 100 percent of attorney hours, the reality is probably somewhere between 5 and 20. Using the Ropes model, 100 hours is about 5 percent of a 1,900-hour target, so if the firm believes that well-oiled AI can boost productivity even 6 percent it’s worth it.

Put aside building and customizing with AI, the proliferation of AI tools requires some experimentation. We’re entering the kitchen sink era when it comes to AI tools. Firms are now willing to buy competing products and let teams and individual lawyers sort out their preferences. Beyond competition, products are announcing MCP connections to complementary products every day, creating a network of warp tubing like a Super Mario Bros. level. Figuring out which user interface the lawyer prefers to do all their work in is an experiment in itself.

For now, these programs represent an exception rather than the rule. But every firm planning to get serious about AI has to come to grips with the fact that this tech isn’t going to optimize itself. Lawyers need time to make it suit their style.

Earlier: The Grace To Dabble: Two Biglaw Firms Look To An AI-First Future
Agentic AI And The QWERTY Problem

he Grace To Dabble: Two Biglaw Firms Look To An AI-First Future

Both firms are investing real resources and forgoing billable hours. They are walking the walk.

By Stephen Embry on November 17, 2025 6:16 pm

Lots of questions swirl around the use of AI by lawyers in general and young lawyers in particular. The questions all center around how do we train young lawyers to use AI tools safely and proficiently while ensuring they develop the skills firms want them to have?

At least two leading firms have recently taken affirmative steps to do just that and prepare for a future where proficiency in the use of AI tools will be critical. Both firms have elected to make an investment in the future that we often don’t see in the legal community. Both firms have elected to forgo some billable hours and perhaps revenue for future returns. Both firms aren’t just talking professional development for young lawyers, they’re investing real resources and forgoing billable hours. They are walking the walk.

The Ropes & Gray Initiative

The first is Ropes & Gray, headquartered in Boston. Ropes has over 1,500 attorneys with 16 offices worldwide and is an Am Law 50 firm.  Definitely qualifies as Biglaw. Ropes has decided to make a sizeable investment in AI: starting immediately, first-year associates can spend up to 400 hours of their annual 1,900-hour billable hour requirement for AI training and experimentation. This can include time experimenting and working in groups. That’s up to 20% of the total requirement. The time, of course, can’t be billed to clients.

According to Amy Ross, Chief of Attorney Talent, and Jane Rogers, Partner & Co-head of the Finance Practice, the associates will work in mentoring circles to discuss how they have used AI, what success they have found and importantly what opportunities they have identified. The associates will also receive hands on training. “Associates are expected to co-create solutions, act as thought partners, and help establish repeatable, defensible approaches for applying AI in practice.” The associates will use Ropes’ 15+ approved AI tools. The firm also plans to “collect analyze and disseminate use cases to enhance learning,” say Ross and Rogers.

It’s a good investment with little downside. On the face of it, it looks like Ropes is giving up a chunk of revenue from the 400 hours that the first-year associates could perhaps bill. I say perhaps because that’s not necessarily true. Let’s face it, first-years are not terribly profitable. They lack the experience to do a whole lot substantially.

In Biglaw, they are highly paid but often don’t produce a lot of revenue. Nor are they expected to. It’s a training period. Add to this the fact that lots of clients are simply unwilling to pay for work by beginning associates. So Ropes’ loss or revenue may not be that significant

But the revenue analysis is not the whole story. Retaining associates these days is an arms race. Keeping associates happy is important:  a firm may spend a year training a first-year associate and not getting much return only to see them leave and lose the investment. My guess is Ropes’ first-year associates will embrace this program, encouraging their long-term and short-term loyalty.

So, all in all, it’s a smart move that will result in AI-first lawyers that are more productive and happier in the long run. And that will inure to the firms’ benefit long term.

The Grace to Dabble

As I recently discussed, mastering GenAI simply requires taking the time work with it, to try and fail. It requires on the job training more than classroom instruction.

I recently did a podcast interview of  Thomas Suh, a former practicing lawyer and founder of LegalMation, an AI litigation tools provider. In his former life as a lawyer, he helped build what he called a tech-first law firm. He and his partners did that by giving themselves the grace to “dabble” as he put it. To master tech by doing what I just said: experimenting, trying different things. Yes, it takes time. Yes, much of that time can’t be billed. But the returns can be significant. Ropes has just given its first years the power to dabble and learn how to leverage AI tools. I’m guessing the return will be significant.

The Latham Program

While Ropes focused on individual experimentation, another Am Law behemoth took a slightly different approach.

A second AI training initiative that will pay off long term is that of Latham & Watkins. Latham is a Los Angeles-based firm and even bigger than Ropes. It has over 3,500 attorneys and over 25 offices worldwide. It consistently ranks near the top of the American Lawyer Am Law revenue rankings. Definitely big, Biglaw.

According to a recent article in Business Insider, Latham took the unusual step in 2024  of  bringing all 400 of its first-year associates to Washington, D.C., for a mandatory two-day AI Academy. The program focused on the use of Harvey and Microsoft Copilot and included outside speakers. Latham repeated the program this year.

In addition to the two-day inaugural program, Latham says there are also multiple events and training programs throughout the year aimed at different groups. The program is “designed to equip lawyers to navigate the significant changes that artificial intelligence (AI) will bring to businesses across industries and around the globe,” according to a Latham announcement.

Like Ropes, Latham views AI as a “generational opportunity” according to one of the Latham partners interviewed for the article. And like Ropes, rather than wringing its hands about what the future will be like, it’s leaning in and making an investment in the future. No doubt those 400 associates left the Academy emboldened to use the AI tools and begin their experimentation as well.

Where Does That Leave Others?

Two Biglaw firms. Two similar approaches. I like the Ropes approach since it encourages experimentation and innovation on an ongoing basis. I like the Latham approach because it sets up the guardrails and expectations.

But either way, approaches like these will position these firms for greater success and help them define what Suh would no doubt call an AI-first approach. Both firms seem to get it: AI is going to fundamentally change the practice of law

Too many firms are doing the opposite: wringing their hands about the future while doing little. Trying to drive forward by looking in the rear-view mirror.

What these two firms are doing is a bit unusual. Investing substantial time and energy in non-billable programs that will aid the firms long term, not just this year. In an industry where partners obsess over utilization rates and billable hour quotas, these firms are making a counterintuitive bet that the short-term sacrifice will yield long-term competitive advantage.

Deal With It

And you don’t have to be an Am Law 50 firm to take the reins. It requires a commitment by firm leadership no matter what the firm size to see and plan for a potentially different future. It requires a senior management that itself is willing to learn. A 50-lawyer firm can’t bring 400 associates to Washington, but it can dedicate Friday afternoons to AI experimentation or partner junior associates with tech-savvy partners. Assuming of course that there are partners willing to become tech savvy.

Moreover, in a world where competition and retention of quality associates becomes more difficult every day, firms that treat AI training as optional professional development are setting themselves up to lose talent to competitors who embrace it as core competency.

Want to compete with Ropes and Latham in years to come?  Stop wondering if AI will disrupt the practice of law and start developing the talent to deal with it.

---

Stephen Embry is a lawyer, speaker, blogger, and writer. He publishes TechLaw Crossroads, a blog devoted to the examination of the tension between technology, the law, and the practice of law.

#

#

TLDR

TLDR: Two articles, a year apart, describe how big law firms are making AI learning part of lawyers’ paid working lives. The distinction is between an intensive introduction for new lawyers and ongoing experimentation across a broader workforce.

When / article	Firm and place	What actually happened
2025 article, describing an academy launched in 2024 and repeated in 2025	Latham & Watkins, described as Los Angeles–based; training held in Washington, D.C.	Brought roughly 400 first-year associates together for a mandatory two-day AI Academy, focused on Harvey and Microsoft Copilot, with additional training throughout the year.
November 2025 article	Ropes & Gray, headquartered in Boston	Allowed first-year associates up to 400 hours of their 1,900-hour annual target for AI training and experimentation. Included mentoring circles, hands-on instruction, and sharing useful approaches across the firm.
October 2026 article	Ropes & Gray, expanding eligibility across its associate ranks	Allowed all associates up to 100 hours toward annual billable targets for AI experimentation. The article presents this as an expansion of the earlier initiative; it does not clearly establish whether first-years retain their separate 400-hour allowance.

These are credits toward internal work targets—not AI training charges sent to clients. Pasted text(20261015-052527)

The useful comparison: Latham’s academy creates a common starting point: tools, expectations, and instruction. Ropes’ protected hours create room for repeated practice on actual workflows. Those approaches could complement each other well: a two-day course gets people started; months of experimentation help them become proficient.

The development across years: The 2025 story emphasizes shaping the incoming generation of lawyers. The 2026 story broadens the effort to existing associates. That suggests AI proficiency is becoming an expectation across the workforce, although the articles don’t establish how consistently lawyers participate.

My MBA take: The strongest idea is the incentive change. Telling associates to innovate while judging them by billable hours makes experimentation personally costly. Giving credit removes that conflict. But both writers are enthusiastic about the investment without showing measured returns—no before-and-after productivity, quality, profitability, or retention data. We have concrete organizational commitments, with their business payoff still unproven in these articles.

##

TLDR 2 (ME)

Yes—I could read the post through the search index, although the direct page wouldn’t load. There’s a stronger connection to the law-firm articles than simply “professionals must learn new technology.”

Your May 2026 post follows your own progression from Index Medicus, handwritten references, and photocopying, through librarian-mediated searches, to PubMed and downloadable articles. You then connect that history to legal research, digital film editing, and word processing. The underlying intellectual work rests on an infrastructure that keeps changing. bqwebpage.blogspot.com

Here are the connections I find most interesting:

1. The eventual benefit may be more ambitious work, rather than the same work done faster.
Consider a hypothetical literature review: easier retrieval might let someone examine 150 relevant papers instead of 30, follow unexpected connections, and revise the question. The project could take just as long while becoming substantially better. Likewise, AI-trained lawyers might investigate more alternatives or test more arguments. “Hours saved” could miss much of the return. That qualifies my earlier emphasis on the articles’ missing productivity measurements.

2. Your film-editing analogy is particularly revealing.
When revision becomes cheap, people can try versions they previously would never have attempted. Applied to legal or consulting work, the benefit could include exploring three competing explanations, restructuring an argument, or asking how an opponent would attack it. Experimentation changes the finished product—not merely its production time.

3. The firms are funding the transition between infrastructures.
Ropes’ protected hours acknowledge that a capable professional can temporarily become slower while learning a new way of working. Latham’s academy supplies an initial shared vocabulary. Your decades-long account makes those initiatives look like deliberate management of a familiar historical transition.

4. There is also a consequential difference.
PubMed mainly changed access to other people’s intellectual work. Generative AI can propose the synthesis, interpretation, and prose itself. That brings the tool much closer to the professional’s judgment. Learning to evaluate its contribution becomes part of learning to use it.

The surprising implication: We should ask not only, “Did those 100 training hours save 100 hours later?” but also, “What work became feasible—and what level of work will clients eventually expect?” Your historical examples suggest that the standard of an adequate professional work product can rise along with the tools.

Vitali Explains My Neuro Price Finding

 Bruce: solid analytical follow-through, wrapped in an ambitious sales pitch. More than padding; less than a breakthrough. They took your observation—“these neuro prices went UP”—and investigated the payment patterns underneath it. They also explicitly credit you.

What have they actually added?

Their contribution	My assessment
Payment distributions: p-tau’s reported 10th–90th percentile spans $25.91–$420; 22.5% of tests were paid below half the median.	Useful original digging. Much more informative than the median alone.
Recurring price points: 345 p-tau tests paid exactly $17.27; other payments cluster at percentages of common reference amounts.	Interesting evidence of inherited fee schedules or formula-based contracting. The explanation is plausible, but the anonymous data cannot establish which contracts produced it.
Timing: private payments were collected while these new codes lacked a national Medicare price.	Probably their strongest strategic insight: early commercial pricing can shape the subsequent Medicare benchmark.
The 98% pattern: 55 codes have medians around 98% of the existing Medicare rate.	An intriguing lead suggesting Medicare Advantage/sequestration feedback. Worth investigating; not proof of the mechanism.

Those are substantive additions reported in the paper. vitali pricing tau pama.pdf

Where the argument gets ahead of the evidence

• Low payment does not establish underpayment. A $25 payment alongside someone else’s $420 payment could reflect different valid contracts. They acknowledge that contracts must be checked, but repeatedly convert price dispersion into a pitch for recoverable revenue. They have not measured how much is actually recoverable.
• Recovering denials does not necessarily improve the PAMA price. Recover previously denied claims at $50, and you could lower a $210 median. Recover them above $210, and you might raise it—or leave it unchanged. Correcting an existing underpayment and adding a newly paid claim are different statistical interventions.
• No national price is an explanation for uncertainty, not proof of why prices rose. Their chronology is persuasive; causation would require more evidence about contracts, payer mix and comparable codes.
• Their recovery statistics do not validate this neuro opportunity. They explicitly say they lack enough neuro claims. The impressive appeal-success percentages come from other procedures and customers, without a comparison showing the incremental benefit of their platform.
• “Same protein” is a good headline, but incomplete economics. It exposes a coding/payment discontinuity; it does not establish that the underlying services, methods and evidence are interchangeable.

Is it true? The basic PAMA framework and overall counts check against CMS: 186 higher medians, 1,171 lower, and these remain preliminary rates. CMS also confirms that final-payment timing matters. I have not independently reproduced their raw-data percentiles and price-cluster calculations, so those remain their reported findings. CMS

The MBA insight I would retain: this illustrates a possible feedback loop—Medicare influences commercial contracts, then those commercial payments reset Medicare. A new code without a national benchmark temporarily loosens that loop. A proprietary code with an unfavorable initial crosswalk may instead carry that low anchor forward. That makes initial coding, pricing and contracting potentially consequential well beyond launch year.

Their commercial purpose is clear: turn your pricing anomaly into a reason to buy Recovr’s claims-recovery service. They demonstrate an interesting pricing problem and sensible audit targets. They do not demonstrate neuro-specific recovered dollars, improved future medians, or superior AI performance. That distinction is the difference between a good market-analysis paper and a demonstrated customer-success case study.