How AI Will Replace Financial Analysts — The OSWorld Threshold Meets the $180 Billion Analyst Economy

The Profession That Trained Wall Street

For sixty years, the financial analyst role has been the default on-ramp into finance. Investment banking analyst programs at Goldman Sachs, Morgan Stanley, JPMorgan, and their peers hired roughly eight thousand new graduates per year into two-year programs that shaped entire careers. Corporate FP&A functions at every public company did the same at a larger scale. The job description — pulling data, building models, writing memos, flagging risks, preparing materials for senior decision-makers — has been so stable that the skills taught to a 2004 Goldman analyst were recognizable to a 2024 one.

That stability is ending. Not because the work is disappearing, but because the human component of the work is being compressed into a narrower and narrower slice of the stack. The capabilities required to execute a financial analyst's day-to-day workflow — pulling data from multiple systems, cross-referencing it against industry sources, building models in Excel and now Python, writing memos that track specific firm-level conventions, preparing decks — are exactly the capabilities that crossed the human baseline on the OSWorld-V benchmark on April 15, 2026. The match between what financial analysts do and what autonomous AI coworkers can now do is not loose. It is close to one-to-one.

The global financial analyst workforce, across investment banks, buy-side firms, corporate FP&A functions, equity research, credit analysis, and consulting, is approximately 4.2 million people worldwide. The aggregate compensation for this workforce is approximately $180 billion per year, with roughly 60 percent of that concentrated in North America and Europe. This is the economic target surface that the autonomous coworker transition is about to point itself at.

What Financial Analyst Work Actually Is

The first step in any serious displacement analysis is to decompose the job into its actual task content. "Financial analyst" is a broad label covering several distinct workflow types, each with different exposure to autonomous AI. The taxonomy matters because the displacement timeline is not uniform.

Category one is data assembly and cleansing. Pulling data from Bloomberg, FactSet, S&P Capital IQ, internal ERP systems, and customer-relationship tools. Normalizing it into consistent formats. Correcting obvious errors. Cross-referencing values between sources when they disagree. This is roughly 30-45 percent of junior analyst time across every category of financial analyst work.

Category two is model building and maintenance. Constructing DCFs, LBO models, merger models, sensitivity tables, budgeting and forecasting models, capital allocation analyses. Updating existing models with new quarters of data. Extending them to handle new scenarios. This is roughly 20-30 percent of junior analyst time and substantially more at the senior-associate level.

Category three is written output. Investment committee memos, credit memos, management decks, board presentations, research notes, variance commentary. Translating numerical outputs into prose that supports a recommendation. This is roughly 15-25 percent of analyst time, heavily weighted toward senior juniors and associates.

Category four is judgment work. Deciding what questions to ask of the data. Noticing anomalies that do not announce themselves. Interpreting ambiguous industry signals. Building relationships with counterparties. Challenging senior decisions when appropriate. Understanding the institutional and political context of a recommendation. This is where the hardest-to-automate value sits, and it is roughly 10-20 percent of analyst time depending on seniority.

The blue bars show the percentage of typical analyst time in each category. The red bars show the percentage of tasks in that category that fall within the capabilities OSWorld-V parity has unlocked. The aggregate weighted exposure — what fraction of actual analyst work is now within autonomous coworker capability — comes out to roughly 72 percent. That number is the headline number for this entire analysis. Seventy-two percent of the work in a typical financial analyst role is now technically substitutable by an AI system that beats the human baseline on the closest benchmark.

The remaining 28 percent is not a small residual. It is concentrated in judgment work, which is durably valuable and which is not going away. But the 72 percent being substitutable means the staffing ratios of analyst teams are about to change dramatically. The question is not whether firms will restructure around this capability. The question is how fast.

The Capability Curve

To understand why the displacement timeline is compressed, you have to look at the capability curve for agentic AI systems over the last three years and project forward.

In 2023, a Claude 2 or GPT-4 could pass the bar exam but could not reliably build a three-statement financial model from source data. It could write convincing prose but could not navigate multi-application workflows. It could summarize a 10-K but could not pull the 10-K from EDGAR, reconcile the numbers against a Bloomberg terminal, and build a comparable-company analysis without human intervention at every step.

By 2024, with the arrival of Claude 3.5 Sonnet and GPT-4o, the capability envelope expanded but still required human orchestration at the workflow level. Analysts who learned to chain these models together got real productivity gains. But the AI was still the tool. The analyst was still the operator.

2025 marked the beginning of the agentic transition. Claude Opus 4, Claude Opus 4.5, and then Opus 4.6 delivered progressively more capable tool use and sustained multi-step execution. The first enterprise deployments of financial-analysis-specific agents appeared in Q3 2025, initially in mid-market firms and specific FP&A functions. They were not replacements. They were high-leverage assistants that compressed a three-day analyst task into four hours.

2026 is the year that ends. OSWorld-V parity means the model is no longer an assistant. It is a functional peer. It can take the same ambiguous memo request a managing director would give a junior analyst, execute the research and modeling autonomously, and produce output that is indistinguishable from — and often better than — what the junior would have produced.

Internal benchmarks from three mid-market buy-side firms that have deployed agentic financial analysis tools in production during Q1 2026 show task-completion rates moving from roughly 56 percent in late 2025 to 83 percent in April 2026. These numbers are above the human-baseline rates for comparable-quality output at those firms. The curve has inflected.

The Displacement Taxonomy

Not every financial analyst role is equally exposed. Decomposing the 4.2 million person workforce into exposure categories reveals a highly uneven displacement landscape.

Investment banking analysts (~28,000 globally at top-tier firms, ~120,000 total including tier-two firms). Highly exposed. Investment banking analyst work is heavily weighted toward data assembly, model construction, and document preparation — the categories with the highest OSWorld-V substitutability. The counterweight is that investment banks have used the analyst program as both a filtering mechanism for senior hires and a client-service theater — neither of which autonomous AI can replicate cleanly. The likely outcome is not the elimination of the analyst program but its radical compression: the size of incoming classes is likely to shrink by 40-60 percent over the next three years, with the remaining analysts doing a different mix of work, more of it relationship- building and agent supervision.

Corporate FP&A analysts (~800,000 globally). Very highly exposed. FP&A work is almost entirely internal, almost entirely document-and-model-based, and lacks the client-facing dimension that partially protects banking analysts. Headcount reductions of 50-70 percent over the next four years are plausible. The survivors will be the ones who transition into "FP&A supervisors" — senior professionals who oversee agent-driven analysis and make the judgment calls the agents cannot.

Equity research and sell-side analysts (~14,000 globally). Moderately exposed. Research work is weighted more heavily toward judgment and thesis-building than transactional analyst work. The thesis-generation function is hard to automate. The supporting infrastructure — data assembly, model maintenance, draft writing — is highly automatable. Expect team sizes to shrink by 30-50 percent while individual senior analyst productivity rises dramatically.

Credit and risk analysts (~400,000 globally). Highly exposed, with a regulatory wrinkle. Credit analysis is almost entirely rules-based pattern recognition on structured data — exactly what autonomous coworkers do well. But credit decisions carry regulatory obligations around documented human judgment that will slow replacement rates even as the economic incentive to replace grows. Expect displacement to lag capability by 12-24 months in regulated markets, which will produce interesting competitive dynamics between banks and less-regulated fintech lenders.

Buy-side analysts at hedge funds, private equity, and asset managers (~180,000 globally). Moderately exposed, with a twist. The thesis-generation and primary-research components are durably human-valuable, but the supporting analytical work is highly automatable. The economics of buy-side employment — high compensation for relatively small headcount — make each individual displacement economically consequential in ways that do not apply to the larger populations.

Consulting financial analysts (McKinsey, Bain, BCG, and tier-two firms, ~65,000 globally). Heavily exposed, especially at the junior level. The consulting analyst job has always been partly a credentialing function and partly a client-deliverable production function. Autonomous coworkers can do the production function. The credentialing function will survive but with radically reduced headcount.

Gray bars are current 2026 headcount in thousands. Red bars are projected 2029 headcount under the base-case displacement scenario. Aggregate headcount across these categories drops from roughly 1.58 million to roughly 780 thousand — a reduction of just over 50 percent across five years. That is the base case. The bull case (slower regulatory environment, slower deployment, stronger client-protection inertia) might see only a 30 percent reduction. The bear case (aggressive enterprise deployment, regulatory neutrality, large-firm competitive pressure) might see closer to 65 percent.

The Snap Signal and the Canary Pattern

On April 15, 2026 — one day before this article was published — Snap Inc. announced plans to lay off up to 16 percent of its global workforce, explicitly citing AI-driven efficiencies as the justification. Snap's cuts were concentrated in three functions: engineering support, customer operations, and — notably — the FP&A and financial operations teams that had grown through the 2023-2025 hiring period.

This is the canary pattern. Before Oracle's 25,000-role reduction, before Amazon's 16,000 corporate layoffs in January, before Block's 40 percent cuts, the signal that finance work was specifically in the crosshairs came through the smaller public-company announcements in late Q1 and early Q2 2026. The larger-cap cuts followed the pattern but obscured the specificity because they were spread across many functions.

What makes financial analyst displacement uniquely legible in the 2026 layoff data is that the function is cleanly identifiable in org charts, the headcount reductions can be quantified against prior quarters' 10-Ks, and the cost savings per displaced FTE are substantial enough to show up in margin improvements that analysts (ironically) are paid to notice.

The yellow bars are the percentage of workforce cut. The red bars are raw FTE counts. The pattern visible across these announcements is that the larger companies are cutting more total bodies but proportionally smaller slices, while the smaller companies are cutting proportionally much larger fractions. That pattern is consistent with the autonomous coworker transition happening first at companies where finance and operations functions are large enough to be consolidated aggressively but not so large that the political cost of consolidation is prohibitive.

Mid-market public companies — $500M to $5B in revenue, with finance teams of 40 to 200 people — are the most exposed population for Q2 and Q3 2026 reductions.

Why This HAR Analysis Is Different From Prior Waves

Previous technology-driven displacement waves in financial analysis — the spreadsheet revolution of the 1980s, the Bloomberg terminal's rollout through the 1990s, the rise of programmatic trading in the 2000s, the data science revolution of the 2010s — all displaced some fraction of the analyst workforce but ultimately expanded the profession rather than contracting it. Each wave created new kinds of analyst work that required the old skills plus new ones. The total headcount grew even as specific sub-specialties shrank.

The autonomous coworker transition is different because it does not create a new category of work that requires human execution. It creates a category of work that can be done without human execution. The prior waves substituted capital for specific labor tasks while still requiring human operators. This wave substitutes capital for the operator function itself.

The historical analogy that matters is not the spreadsheet. It is the ATM. When ATMs were introduced in the 1970s and 80s, many predicted they would eliminate bank tellers. What happened instead is that ATMs eliminated the transactional component of the teller job, but bank branches responded by redeploying tellers into relationship and sales roles. Total teller headcount barely fell, and in fact rose for two decades as banks expanded branch networks.

Autonomous coworkers are the ATM pattern only if financial analysts have a parallel redeployment option — a relationship and judgment function that humans retain uncompetitively. Some analysts will have that option. Most will not, because the relationship function at investment banks and FP&A teams is already concentrated in managing directors and vice presidents, not in the junior analyst population that represents 70 percent of the workforce.

The headcount trajectory. Steady growth through four decades of technology adoption, then a sharp reversal starting in 2025. The inflection is the single most important shift in the profession in sixty years.

Who Survives

The survival map for financial analysts over the next five years looks like this.

The deep judgment specialists. Analysts who cover industries or asset classes where the value is in non-obvious thesis generation, not in data processing. Distressed debt specialists. Emerging- markets research analysts. Specialist equity research in sectors where primary research is irreplaceable. These roles are roughly 5-8 percent of the current analyst workforce and will grow in compensation even as the total population shrinks.

The agent supervisors. Analysts who master the skill of directing, validating, and trusting autonomous agent output. This is a genuine new skill — not a trivial extension of existing analyst skills. It requires understanding model failure modes, knowing how to write specifications that produce reliable output, and knowing when to trust an agent's conclusion versus when to override it. The firms that train analysts in this discipline well will capture the economic surplus. This role is approximately 15-25 percent of the current workforce and will shrink in absolute headcount while commanding substantial compensation premiums.

The client-facing relationship specialists. Analysts whose primary function is representing the firm to external counterparties — managing relationships with portfolio company CFOs, sell-side analysts covering portfolio positions, limited partners at private funds, and similar roles. This is approximately 10 percent of the current workforce and is durably human.

Everyone else. The 65-70 percent of the workforce that is doing transactional data, model, and document work that is about to be automated. This population's primary survival strategy is transitioning into one of the three categories above — or out of the profession entirely.

The firms that will produce the most dislocation are the ones that handle this transition poorly. The firms that will produce the most value are the ones that invest aggressively in moving their analyst population into the supervisor and specialist categories before the substitution curve catches up with them.

What the Junior Analyst Career Looks Like Now

For someone starting a financial analyst career in 2026, the job description is about to shift dramatically.

The 2024 junior analyst job was: execute 40-80 hour weeks of data pulling, model building, and document preparation, under the supervision of more senior colleagues, for two years, after which you either exit to the buy side, exit to corporate, or pursue a longer track at the firm.

The 2026 junior analyst job is starting to look more like: co-execute those same 40-80 hour weeks alongside a team of autonomous agents, spending substantial portions of that time validating agent output, correcting agent errors, specifying agent workflows more precisely, and producing the judgment layer that agents cannot produce. The total hours remain similar. The composition of those hours is fundamentally different.

This is either a richer job or a hollowed-out one, depending on the firm's investment in the training that makes the new composition meaningful. At firms that invest in agent-supervision training, the junior analyst role becomes more intellectually demanding, with more of the substantive judgment concentrated into fewer hours. At firms that do not invest, the junior analyst role becomes a kind of quality-assurance position for autonomous systems, with most of the economic value captured by the agents and most of the human time spent on low-leverage validation.

The single most important career question for anyone entering or early in a financial analyst career is: is your firm investing in you as a future supervisor, or using you as a validation cost center?

The Implications For Everyone Else

Even if you are not a financial analyst, the displacement of this profession has consequences that reach into the rest of the economy.

First, the financial analyst workforce has been a default career destination for a specific population of graduates — finance, economics, and accounting majors from selective universities, aggregated through the banking and FP&A on-ramps. If that destination shrinks by 50 percent, the career decisions of roughly 80,000 new US graduates per year have to be rerouted. This affects undergraduate major choices, business school enrollment, and the downstream competitive dynamics of every adjacent profession that historically competed with finance for the same talent.

Second, the cost structure of financial services is about to change meaningfully. The wage bill for financial analyst work in the US alone is roughly $45 billion per year. If 30 percent of that is automated out by 2029 — a modest projection — approximately $13-15 billion per year flows out of wages and into margin, infrastructure spending on AI tools, and competitive price compression. That displaced cash will reshape the profitability and strategic positioning of every major financial services firm.

Third, the regulatory and professional-standards response is going to be messy. Financial analysis has regulatory dimensions — fiduciary duties, auditing standards, reasonable-care obligations — that do not cleanly transfer to autonomous systems. Expect substantial litigation, rule-making, and professional-association activity over the next three to five years as the transition forces a reconsideration of what financial analysis is as a regulated activity.

Fourth, and this is the dimension most analysts are not yet thinking about, the displacement of financial analyst work accelerates the displacement of adjacent knowledge-work professions that depend on analyst-produced outputs. Investment committee members, corporate CFOs, private equity principals, and similar senior roles spend substantial time reviewing analyst output. When that output becomes faster, cheaper, and more numerous, the review function itself changes — and in some cases contracts. The displacement cascade does not stop at the analyst level.

The 2029 headcount map. Approximately half of current positions displaced. Roughly a quarter retained in modified form. The remainder split between new supervisor roles (the agent-oversight function) and new specialist roles (the judgment concentration). The total headcount drops substantially, but the composition shifts toward higher-skilled, higher-compensated work — for those who make the transition.

What To Do If You Are Currently A Financial Analyst

If you are reading this article and you are currently a financial analyst — at an investment bank, a corporate FP&A team, a research firm, a consulting firm, a hedge fund, or anywhere else in the 4.2 million person workforce — the question is what to do in the next six to eighteen months. Three concrete actions matter.

First, audit your own task mix honestly against the OSWorld-V capability list. Track for two weeks what you actually spend your working hours on. Categorize each hour into data assembly, model building, written output, or judgment. The percentage of your time that falls outside the judgment category is your substitutability percentage. If it is above 70 percent, you are in the most exposed bucket. If it is below 40 percent, you have more time than most people in your cohort.

Second, aggressively develop agent supervision skills. This is a specific, teachable discipline. It involves learning to write effective specifications for agent work, understanding how to validate agent output efficiently, knowing which failure modes to watch for, and becoming fluent in at least one major agent platform (Claude Cowork, ChatGPT Enterprise, and Microsoft Copilot Enterprise are the three dominant platforms in finance as of mid- 2026). This skill is not yet taught in MBA programs. The people who develop it independently are going to have a large first-mover advantage.

Third, move toward the judgment-heavy components of your role aggressively, even if it means taking unusual assignments. If you are a junior analyst, volunteer for the complex, high-context assignments that require understanding industry dynamics, firm history, or client relationships. These are the assignments that build the judgment muscle that durably survives the transition. The analysts who spent 2023-2025 building pure execution muscle are in the worst position. The analysts who spent those years building judgment and relationship muscle are in the best position.

For analysts who conclude, after honest self-audit, that their role is in the most exposed category and their firm is not investing in their transition to a supervisor or specialist path, the difficult conversation is about when to leave versus when to stay. The window for voluntary career transitions — with severance, with runway, with upside from the final compensation years before the cuts accelerate — is closing. Analysts who wait until 2027 to make a move will be competing with a large population of similarly situated peers for a smaller pool of alternatives.

The Firm-By-Firm Strategy Map

Not every financial services firm is approaching this transition the same way. A rough taxonomy of how the major categories of employer are positioning themselves is useful for anyone trying to understand which firms offer durable career paths and which are setting themselves up for the kind of reactive, brutal cost-cutting that Snap just executed.

The bulge-bracket investment banks (Goldman Sachs, Morgan Stanley, JPMorgan, Bank of America, Citi). These firms have the most institutional capacity to handle the transition well. They have large training budgets, sophisticated technology teams, and the political capital to restructure analyst programs without immediate revenue loss. Goldman Sachs announced a restructured analyst program in March 2026 that formally includes "agent partnership training" as a first-year curriculum component. JPMorgan has been quieter but has deployed internal tooling to every analyst since late 2025. The likely outcome for this segment is a 40-50 percent reduction in incoming analyst class sizes by 2028, with the retained population heavily trained in supervisor and specialist skills, and with significantly compressed work hours per head because of the agent leverage. These are arguably the best firms to start a career at if you can land there, because the transition is being managed deliberately.

The tier-two investment banks and boutiques (Jefferies, Evercore, Lazard, Houlihan Lokey, and similar). These firms face more acute cost pressure because their margins are thinner and their ability to subsidize transition costs is lower. Expect more abrupt restructuring, with specific teams absorbing the brunt of changes. Boutique advisory shops that built their 2020-2025 growth on aggressive hiring are now in the most vulnerable position — those hires were made assuming sustained revenue growth that the compressed-analyst-headcount era will not deliver.

Private equity and alternative asset managers. Mixed. The largest firms (Blackstone, KKR, Apollo, Carlyle) have been investing in AI-augmented deal teams since 2024 and are relatively well prepared. The mid-market PE firms and the single-strategy hedge funds are more exposed because they have smaller technology teams and less institutional slack for training-heavy transitions. The net effect in PE is likely to be fewer associates doing more substantive work at the largest firms, while mid-market firms consolidate or sell to the larger ones.

Corporate FP&A teams at Fortune 500 companies. Highly exposed and mostly unprepared. Corporate FP&A teams have historically not been at the forefront of technology adoption, and the people running them came up through an era when the main innovation was moving from Excel to Anaplan or similar planning tools. The autonomous coworker transition is categorically different and requires strategic capability that most corporate FP&A leaders have not built. Expect substantial headcount reductions here driven by CFO directives rather than careful transition planning — which is the worst possible way to execute this kind of change.

Consulting firms (McKinsey, Bain, BCG, the Big Four). Highly strategic, highly visible, and highly political. The analyst and associate ranks at these firms have been the training ground for generations of business leaders. The client-value proposition of consulting has always been "we bring smart people who do a lot of work quickly." When autonomous coworkers can do the quick-work portion, the consulting value proposition compresses toward "we bring smart people who have judgment." That is a different business. McKinsey has been publicly quiet about how it is restructuring but has privately been shrinking new-analyst class sizes for two years. Bain and BCG have been more experimental, testing agent-partnered project teams. The Big Four firms face the largest displacement pressure because their analyst populations are largest and their compensation scaling is most sensitive to cost pressure from clients who now have agent alternatives.

Green bars are my qualitative readiness scores for handling the transition well. Red bars are my qualitative risk-of-disruption scores for how much the firm type stands to lose if handled poorly. Firm types with high readiness and low disruption risk (bulge-bracket banks, large alternatives) are the best-positioned employers. Firm types with low readiness and high disruption risk (corporate FP&A, mid-market PE) are the worst positions to be in professionally during the next three years.

The Labor Market Ripple Effects

The displacement of financial analyst work is happening in a labor market that is still processing the prior three years of tech-sector layoffs. The Q1 2026 layoff data — 78,557 total tech layoffs, with roughly 48 percent AI-attributed — represents a labor market that is already struggling to absorb displaced white-collar workers. The addition of meaningful numbers of displaced financial analysts to that pool will test the absorption capacity of the US economy in ways that have not been tested before.

The economic argument that displaced workers will find new, more productive work is historically correct but has specific conditions that may not hold in this case. Those conditions include: a labor market flexible enough to absorb sectoral shifts, a retraining infrastructure capable of moving workers between sectors, and a macroeconomic environment supportive of new hiring in the sectors that expand. The US labor market in 2026 has some of these conditions and not others. Retraining infrastructure is particularly weak. The sectors that have been expanding — AI infrastructure, specialized cybersecurity, regulated healthcare — require specific credentials and skills that do not transfer easily from financial analysis.

Expect, as a result, a lengthened friction period between displacement and reabsorption. Historical displacement cycles have run 18-36 months between peak layoffs and restabilized employment in the affected population. The autonomous coworker cycle may run longer because the displacement is happening in professions whose skills transfer poorly to the growth sectors. Two to four years of structural unemployment in the affected analyst population is the base case, which translates into political pressure, policy response, and likely specific regulatory and retraining interventions that will further shape the transition.

The End of the Default Career

The financial analyst role has been a default career for a specific slice of educated workers for sixty years. It was stable, well- compensated, socially legible, and had a clear progression path. It produced generations of corporate leaders. It trained the most influential people in American finance.

It is ending, or at least contracting sharply enough that the default-career status can no longer be taken for granted. OSWorld-V parity is the technical marker of that end. The Snap layoffs, the Oracle reductions, the Block cuts, and the Q2 2026 FP&A consolidations that will follow are the economic and cultural markers.

What comes next is not the end of financial analysis as a discipline. The discipline will survive. It will be practiced at higher levels, with more leverage, by fewer people. The judgment function at the core of the discipline will actually grow in importance, because the value of judgment increases when execution becomes a commodity.

But the on-ramp is narrowing. The jobs that once served as the training ground for the profession are being automated before the next generation of trainees arrives. The profession's reproduction cycle is being interrupted. How the industry solves that problem — whether by investing aggressively in the next generation of supervisor-analysts, by restructuring training programs around agent partnership, or by simply letting the profession contract until the remaining seats are held by people who came up through the last decade of pre-agentic work — is the most important question facing financial services in the second half of the 2020s.

Everyone inside the profession has an interest in getting the answer right. Most of them do not yet know the question is being asked.

The analysts reading this article in April 2026 are in a specific kind of historical moment. The profession that trained them is changing underneath them, and the version of the career they signed up for is not the career that will be available to the next cohort behind them. That is not, by itself, a tragedy. Every generation of professional work faces a version of this discontinuity. But it is a moment that rewards clear-eyed assessment and penalizes denial.

The work is not disappearing. The work is being redistributed between humans and agents, and the distribution is shifting faster than the institutional mechanisms for training and career development can adapt. Individual analysts who get ahead of that curve — who see the shift, who invest in the skills that remain durably valuable, who move toward the judgment-heavy and supervisor-heavy work — will come out of this transition with stronger careers than the ones they had when they started it.

Individual analysts who wait for their firms to tell them what to do will come out of it with the careers their firms decide to give them, which in most cases will be shorter and smaller than what they expected. The agency belongs to the person willing to read the trajectory and act on it.

The shift has already started. The first Snap cuts are the first signal. The Oracle and Block announcements are not far behind. The next twelve months will define which analysts are still working in this profession five years from now, and which are writing retrospectives about the career they used to have.

It is a transition worth taking seriously, starting now.