Everyone keeps saying it's an employer's market. They're half right.

For generalist roles, yes—companies have leverage. According to Deloitte, 72% of companies cite talent shortages as a major challenge, but that hasn't stopped them from being selective. McKinsey's global survey found 87% of executives say their organizations either face skill gaps already or expect them within five years—yet only 28% are planning to invest in upskilling programs to close them.

But in AI and crypto? Different story. The a16z State of Crypto 2025 report found that while roughly 1,000 workers left crypto for AI startups since ChatGPT launched, the industry gained the same number from traditional finance, tech, and consulting. Talent is flowing in, not out. And the Microsoft/LinkedIn Work Trend Index reports that 66% of leaders say they wouldn't hire someone without AI skills—but the supply isn't there. McKinsey found that fewer than half of potential candidates have the high-demand tech skills listed in job postings.

Two markets. Two sets of rules. Most founders are operating with the wrong mental model for the talent market they're actually in.

As Head of Talent at Lazer Technologies—a consulting firm working across Crypto, AI, and Commerce—I see this confusion play out across hundreds of hiring conversations. Founders apply employer's-market tactics to candidate's-market roles. They move slowly when they should move fast. They filter aggressively when they should be expanding their aperture.

Here are the three shifts that matter most heading into 2026.

The rise of "996 culture" in tech tells you something important about how founders are thinking about the talent market—and where the nuance gets lost.

996—working 9am to 9pm, six days a week—originated in China's tech industry and was declared illegal there in 2021. But it's now showing up as a hiring filter at U.S. startups. AI startup Rilla, based in New York, tells candidates in job listings not to bother applying unless they're excited about working 70+ hours per week in person. Multiple recruiters report that clients require candidates to confirm willingness to work 996 before interviews even begin.

996 might work for where some teams are at. Early-stage, all-hands-on-deck, shipping v1—there's a time when intensity is the strategy. But most founders aren't thinking carefully enough about what talent they need for specific roles, and how that maps to the culture they're building.

Consider: you need a senior ML engineer to optimize your inference pipeline and reduce latency on your LLM serving infrastructure. This is highly specialized work—distributed training, GPU optimization, model quantization. With that experience comes efficiency. A senior engineer who's done this before can architect a solution in days that a junior team might take months to figure out. They produce faster and at higher quality working 45 hours than an inexperienced team grinding 70. But that senior talent has options—and they're often at a life stage where 996 isn't appealing. If you filter them out with an intensity requirement, you might end up with a less experienced team working harder but delivering less.

The market bifurcation makes this more acute. According to Deloitte, nearly 90% of tech industry leaders say recruiting and retaining tech talent remains a moderate or major issue—even during layoff periods. McKinsey found that 92% of companies plan to increase their AI investments over the next three years, but only 1% describe their current AI deployment as "mature." The demand is accelerating. The supply isn't keeping up.

AI and protocol engineers will work hard—that's not the issue. Specialized talent isn't averse to intensity; they just don't need to accept it. They have options. So it becomes about two things: finding the talent that genuinely wants to build with you, and creating the incentive structures that make the grueling work worth it. Equity, ownership, mission, technical challenge—these matter more than mandating hours.

The best teams work hard strategically, then recover intentionally. Sprint during launches, recharge to maintain peak performance. The founders who treat every week like a sprint will burn out their best people. The founders who never ask for a push will get outpaced.

The 996 framing presents a false binary. The reality: know when to flex, know when to pull back, and be transparent with your team about which mode you're in.

What to do about it:

• Audit your open roles: which are in an employer's market, which are candidate-driven? Calibrate your speed, selectivity, and pitch accordingly.

• For candidate-driven roles, compress your interview process to under three weeks.

• Before requiring extreme intensity, ask: does this role actually need 996, or would a more experienced hire working normal hours deliver more?

• Create incentive structures that make intensity worth it—meaningful equity, real ownership, technical challenge. Mandating hours without aligned incentives breeds resentment.

In June, Stripe acquired Privy—a crypto wallet infrastructure company supporting 75 million accounts across 1,000+ developer teams. Stripe CEO Patrick Collison said the goal is to connect "Privy's wallets to the money movement capabilities in Stripe and Bridge" to enable "a new generation of global, internet-native financial services."

That's the CEO of a $91.5 billion payments company describing crypto infrastructure as core to their future. Not adjacent. Core.

The numbers are staggering:

• Coinbase acquired Deribit for $2.9 billion in May 2025—$700 million cash plus 11 million shares—gaining the world's largest crypto options exchange with $30 billion in open interest and over $1 trillion in annual trading volume.

• Kraken bought NinjaTrader for $1.5 billion in March 2025, described as "the largest-ever deal combining traditional finance and crypto."

• Ripple acquired Hidden Road for $1.25 billion in April 2025, becoming the first crypto company to own and operate a global multi-asset prime broker clearing $3 trillion annually.

• Stripe acquired Bridge for $1.1 billion in late 2024, Stripe's largest acquisition to date.

This is the maturation moment. The talent profile that built crypto isn't the same profile you need to scale it.

The a16z State of Crypto 2025 report captures this shift: the industry's talent pipeline is diversifying beyond developers, with growing demand for compliance, infrastructure, and product specialists. Workers entering crypto now tend to come from traditional finance and consulting backgrounds—a sign that the line between TradFi and crypto is blurring.

Electric Capital's 2024 Developer Report shows the technical evolution too. One-third of crypto developers now work across multiple blockchains, up from less than 10% in 2015. The work has become about orchestrating complexity across ecosystems.

What enterprise crypto needs:

Engineers with deep distributed systems experience—people who understand consensus at scale, not just in theory. Strong CS fundamentals matter more than ever, particularly around security: smart contract auditing, cryptographic primitives, secure key management. These aren't skills you pick up casually. A bad hire in a security-critical role can cost a protocol millions.

What crypto-native talent brings:

You can't build enterprise crypto with TradFi talent alone. You need people who understand crypto primitives—how wallets work, how on-chain governance functions, how tokenomics shape incentives. You need builders comfortable with modern frameworks and languages (Rust, Move, Solidity), who've shipped in public, who understand the culture of building in the open. The innovation mindset that characterized early crypto is still essential—it just needs to be paired with operational rigor.

You need both. The scrappy crypto-native builder who understands the ecosystem AND the distributed systems engineer who's scaled infrastructure at a fintech or cloud company.

At Lazer, we've proven this out. One of our highest-impact hires had barely touched crypto—he'd used stablecoins a bit but didn't have real web3 experience. What he had: proven experience building and scaling consumer products. He'd driven around New Zealand testing his app with farmers who were the actual users. He founded his own React Native developer tooling. In our interviews, he demonstrated strong technical problem-solving and the ability to pick up new domains quickly. Now he builds mobile apps for some of the biggest names in the industry.

He didn't have the crypto pedigree. He had the engineering fundamentals and the ownership mindset. The domain knowledge came fast.

What to do about it:

• Stop requiring "crypto experience" for roles where it doesn't matter. Your product manager doesn't need to understand MEV to understand users.

• For security-critical roles, don't compromise on CS fundamentals. Distributed systems experience, cryptography knowledge, and security mindset are non-negotiable.

• When evaluating candidates from TradFi or tech, ask: "Have they scaled something to real users? Can they learn new domains quickly? Do they have the ownership mindset?"

• For crypto-native hires, evaluate their understanding of primitives—not just years in the space. Someone who deeply understands how different L2s work is more valuable than someone who's been around since 2017 but can't explain rollups.

The companies that gutted teams expecting AI to fill the gaps are about to learn an expensive lesson.

McKinsey's January 2025 research found that while nearly all employees and C-suite leaders have some familiarity with gen AI tools, nearly half of employees want more formal training. The gap isn't in awareness—it's in capability. And according to the Microsoft/LinkedIn Work Trend Index, 66% of leaders wouldn't hire someone without AI skills, but only 39% of users have received AI training from their company.

The demand for AI-capable talent is outstripping supply. McKinsey's analysis of 4.3 million job postings found fewer than half of potential candidates have the high-demand tech skills listed. The World Economic Forum estimates that nearly six in ten workers will require training before 2030.

This is where credentials start to break down. The Stanford grad with two years at a big lab might be perfect for frontier research. For building AI products—taking models and shipping things people actually use—the person who built and scaled three apps at a startup no one's heard of might be better. But you'll never know if they don't make it past your ATS.

The market is catching up to this reality. According to McKinsey's Workforce Transformation Report, the percentage of companies adopting skills-based hiring practices increased from 40% in 2020 to 60% in 2024. Deloitte found that 72% of companies cite talent shortages as a major challenge, driving them to explore alternative hiring strategies beyond traditional degree requirements.

This isn't about lowering the bar. It's about measuring the right things.

That includes AI fluency—and most companies still aren't evaluating it.

According to HR Dive, "Hiring someone today without assessing their AI skills is like hiring someone in the 1990s without checking if they could use the internet." Yet most interview processes were built before AI tools became ubiquitous.

The skills to evaluate aren't monolithic. Prompt engineering—do they know how to ask AI the right questions? Critical thinking—can they evaluate AI-generated outputs, spot flaws, and refine them? AI-human balance—do they know when to use AI for efficiency and when to apply human judgment?

Platforms like HackerRank now offer AI-assisted IDE environments where interviewers can observe how candidates collaborate with AI tools in real time, with all conversations captured in detailed reports. This reveals something traditional coding interviews miss: not just whether someone can solve a problem, but how they think about when to leverage AI and when to rely on their own judgment.

At Lazer, our bar is high. Every candidate goes through rigorous technical interviews that assess CS fundamentals and ability to build at scale. But we've learned that the most impactful hires often don't fit the typical mold.

One example: an engineer most companies would've passed on. Younger, no brand names, wasn't actively working in crypto at the time. He was writing AI research papers and dabbling in crypto while playing with AI tools. But in our technical interviews, his CS fundamentals were rock solid. And he'd run a web3 studio from his college dorm room—clear entrepreneurial mindset and bias to action. We hired him. He's now making major contributions to one of the biggest apps in the industry.

He didn't have the pedigree. He had proof of work.

What to do about it:

• Before adding any credential filter to a job description, ask: does this role actually require this? Challenge "3 years crypto experience" or "top-tier CS degree" as defaults.

• Change your interview questions. Ask "What have you built that nobody asked you to build?" and "Walk me through a time you learned a new technology quickly and shipped with it."

• Build AI fluency evaluation into every role. Ask candidates: "Walk me through how you use AI tools in your current workflow. When do you use them? When do you not?"

• For technical roles, consider assessment environments that let you observe AI collaboration in real time—how candidates prompt, evaluate outputs, and integrate AI with their own judgment.

• Track the correlation between your credential filters and actual performance. Most companies have never done this analysis. When they do, they find the correlation is weaker than they assumed.

The market is bifurcated. It's an employer's market for generalist roles and a candidate's market for specialized AI and crypto talent. Most founders are applying the wrong market's rules to their hiring. Know which market you're competing in—and remember that today's leverage can become tomorrow's attrition if you don't calibrate intensity intelligently.

Crypto is going enterprise. The M&A numbers tell the story: $2.9B, $1.5B, $1.25B, $1.1B. You need both crypto-native builders who understand primitives and innovation AND engineers with distributed systems depth, strong CS fundamentals, and security expertise. The domain knowledge can be taught. The engineering foundation can't be faked.

Proof of work is beating credentials. McKinsey found 87% of companies face skills gaps—and skills-based hiring adoption has jumped from 40% to 60% since 2020. The ability to ship, learn fast, and leverage AI tools matters more than the logos on a resume. Evaluate accordingly.

Your competitors are still hiring for the market they think they're in. The best founders are hiring for the market that's coming.

Everyone keeps saying it's an employer's market. They're half right.

For generalist roles, yes—companies have leverage. According to Deloitte, 72% of companies cite talent shortages as a major challenge, but that hasn't stopped them from being selective. McKinsey's global survey found 87% of executives say their organizations either face skill gaps already or expect them within five years—yet only 28% are planning to invest in upskilling programs to close them.

But in AI and crypto? Different story. The a16z State of Crypto 2025 report found that while roughly 1,000 workers left crypto for AI startups since ChatGPT launched, the industry gained the same number from traditional finance, tech, and consulting. Talent is flowing in, not out. And the Microsoft/LinkedIn Work Trend Index reports that 66% of leaders say they wouldn't hire someone without AI skills—but the supply isn't there. McKinsey found that fewer than half of potential candidates have the high-demand tech skills listed in job postings.

Two markets. Two sets of rules. Most founders are operating with the wrong mental model for the talent market they're actually in.

As Head of Talent at Lazer Technologies—a consulting firm working across Crypto, AI, and Commerce—I see this confusion play out across hundreds of hiring conversations. Founders apply employer's-market tactics to candidate's-market roles. They move slowly when they should move fast. They filter aggressively when they should be expanding their aperture.

Here are the three shifts that matter most heading into 2026.

The rise of "996 culture" in tech tells you something important about how founders are thinking about the talent market—and where the nuance gets lost.

996—working 9am to 9pm, six days a week—originated in China's tech industry and was declared illegal there in 2021. But it's now showing up as a hiring filter at U.S. startups. AI startup Rilla, based in New York, tells candidates in job listings not to bother applying unless they're excited about working 70+ hours per week in person. Multiple recruiters report that clients require candidates to confirm willingness to work 996 before interviews even begin.

996 might work for where some teams are at. Early-stage, all-hands-on-deck, shipping v1—there's a time when intensity is the strategy. But most founders aren't thinking carefully enough about what talent they need for specific roles, and how that maps to the culture they're building.

Consider: you need a senior ML engineer to optimize your inference pipeline and reduce latency on your LLM serving infrastructure. This is highly specialized work—distributed training, GPU optimization, model quantization. With that experience comes efficiency. A senior engineer who's done this before can architect a solution in days that a junior team might take months to figure out. They produce faster and at higher quality working 45 hours than an inexperienced team grinding 70. But that senior talent has options—and they're often at a life stage where 996 isn't appealing. If you filter them out with an intensity requirement, you might end up with a less experienced team working harder but delivering less.

The market bifurcation makes this more acute. According to Deloitte, nearly 90% of tech industry leaders say recruiting and retaining tech talent remains a moderate or major issue—even during layoff periods. McKinsey found that 92% of companies plan to increase their AI investments over the next three years, but only 1% describe their current AI deployment as "mature." The demand is accelerating. The supply isn't keeping up.

AI and protocol engineers will work hard—that's not the issue. Specialized talent isn't averse to intensity; they just don't need to accept it. They have options. So it becomes about two things: finding the talent that genuinely wants to build with you, and creating the incentive structures that make the grueling work worth it. Equity, ownership, mission, technical challenge—these matter more than mandating hours.

The best teams work hard strategically, then recover intentionally. Sprint during launches, recharge to maintain peak performance. The founders who treat every week like a sprint will burn out their best people. The founders who never ask for a push will get outpaced.

The 996 framing presents a false binary. The reality: know when to flex, know when to pull back, and be transparent with your team about which mode you're in.

What to do about it:

• Audit your open roles: which are in an employer's market, which are candidate-driven? Calibrate your speed, selectivity, and pitch accordingly.

• For candidate-driven roles, compress your interview process to under three weeks.

• Before requiring extreme intensity, ask: does this role actually need 996, or would a more experienced hire working normal hours deliver more?

• Create incentive structures that make intensity worth it—meaningful equity, real ownership, technical challenge. Mandating hours without aligned incentives breeds resentment.

In June, Stripe acquired Privy—a crypto wallet infrastructure company supporting 75 million accounts across 1,000+ developer teams. Stripe CEO Patrick Collison said the goal is to connect "Privy's wallets to the money movement capabilities in Stripe and Bridge" to enable "a new generation of global, internet-native financial services."

That's the CEO of a $91.5 billion payments company describing crypto infrastructure as core to their future. Not adjacent. Core.

The numbers are staggering:

• Coinbase acquired Deribit for $2.9 billion in May 2025—$700 million cash plus 11 million shares—gaining the world's largest crypto options exchange with $30 billion in open interest and over $1 trillion in annual trading volume.

• Kraken bought NinjaTrader for $1.5 billion in March 2025, described as "the largest-ever deal combining traditional finance and crypto."

• Ripple acquired Hidden Road for $1.25 billion in April 2025, becoming the first crypto company to own and operate a global multi-asset prime broker clearing $3 trillion annually.

• Stripe acquired Bridge for $1.1 billion in late 2024, Stripe's largest acquisition to date.

This is the maturation moment. The talent profile that built crypto isn't the same profile you need to scale it.

The a16z State of Crypto 2025 report captures this shift: the industry's talent pipeline is diversifying beyond developers, with growing demand for compliance, infrastructure, and product specialists. Workers entering crypto now tend to come from traditional finance and consulting backgrounds—a sign that the line between TradFi and crypto is blurring.

Electric Capital's 2024 Developer Report shows the technical evolution too. One-third of crypto developers now work across multiple blockchains, up from less than 10% in 2015. The work has become about orchestrating complexity across ecosystems.

What enterprise crypto needs:

Engineers with deep distributed systems experience—people who understand consensus at scale, not just in theory. Strong CS fundamentals matter more than ever, particularly around security: smart contract auditing, cryptographic primitives, secure key management. These aren't skills you pick up casually. A bad hire in a security-critical role can cost a protocol millions.

What crypto-native talent brings:

You can't build enterprise crypto with TradFi talent alone. You need people who understand crypto primitives—how wallets work, how on-chain governance functions, how tokenomics shape incentives. You need builders comfortable with modern frameworks and languages (Rust, Move, Solidity), who've shipped in public, who understand the culture of building in the open. The innovation mindset that characterized early crypto is still essential—it just needs to be paired with operational rigor.

You need both. The scrappy crypto-native builder who understands the ecosystem AND the distributed systems engineer who's scaled infrastructure at a fintech or cloud company.

At Lazer, we've proven this out. One of our highest-impact hires had barely touched crypto—he'd used stablecoins a bit but didn't have real web3 experience. What he had: proven experience building and scaling consumer products. He'd driven around New Zealand testing his app with farmers who were the actual users. He founded his own React Native developer tooling. In our interviews, he demonstrated strong technical problem-solving and the ability to pick up new domains quickly. Now he builds mobile apps for some of the biggest names in the industry.

He didn't have the crypto pedigree. He had the engineering fundamentals and the ownership mindset. The domain knowledge came fast.

What to do about it:

• Stop requiring "crypto experience" for roles where it doesn't matter. Your product manager doesn't need to understand MEV to understand users.

• For security-critical roles, don't compromise on CS fundamentals. Distributed systems experience, cryptography knowledge, and security mindset are non-negotiable.

• When evaluating candidates from TradFi or tech, ask: "Have they scaled something to real users? Can they learn new domains quickly? Do they have the ownership mindset?"

• For crypto-native hires, evaluate their understanding of primitives—not just years in the space. Someone who deeply understands how different L2s work is more valuable than someone who's been around since 2017 but can't explain rollups.

The companies that gutted teams expecting AI to fill the gaps are about to learn an expensive lesson.

McKinsey's January 2025 research found that while nearly all employees and C-suite leaders have some familiarity with gen AI tools, nearly half of employees want more formal training. The gap isn't in awareness—it's in capability. And according to the Microsoft/LinkedIn Work Trend Index, 66% of leaders wouldn't hire someone without AI skills, but only 39% of users have received AI training from their company.

The demand for AI-capable talent is outstripping supply. McKinsey's analysis of 4.3 million job postings found fewer than half of potential candidates have the high-demand tech skills listed. The World Economic Forum estimates that nearly six in ten workers will require training before 2030.

This is where credentials start to break down. The Stanford grad with two years at a big lab might be perfect for frontier research. For building AI products—taking models and shipping things people actually use—the person who built and scaled three apps at a startup no one's heard of might be better. But you'll never know if they don't make it past your ATS.

The market is catching up to this reality. According to McKinsey's Workforce Transformation Report, the percentage of companies adopting skills-based hiring practices increased from 40% in 2020 to 60% in 2024. Deloitte found that 72% of companies cite talent shortages as a major challenge, driving them to explore alternative hiring strategies beyond traditional degree requirements.

This isn't about lowering the bar. It's about measuring the right things.

That includes AI fluency—and most companies still aren't evaluating it.

According to HR Dive, "Hiring someone today without assessing their AI skills is like hiring someone in the 1990s without checking if they could use the internet." Yet most interview processes were built before AI tools became ubiquitous.

The skills to evaluate aren't monolithic. Prompt engineering—do they know how to ask AI the right questions? Critical thinking—can they evaluate AI-generated outputs, spot flaws, and refine them? AI-human balance—do they know when to use AI for efficiency and when to apply human judgment?

Platforms like HackerRank now offer AI-assisted IDE environments where interviewers can observe how candidates collaborate with AI tools in real time, with all conversations captured in detailed reports. This reveals something traditional coding interviews miss: not just whether someone can solve a problem, but how they think about when to leverage AI and when to rely on their own judgment.

At Lazer, our bar is high. Every candidate goes through rigorous technical interviews that assess CS fundamentals and ability to build at scale. But we've learned that the most impactful hires often don't fit the typical mold.

One example: an engineer most companies would've passed on. Younger, no brand names, wasn't actively working in crypto at the time. He was writing AI research papers and dabbling in crypto while playing with AI tools. But in our technical interviews, his CS fundamentals were rock solid. And he'd run a web3 studio from his college dorm room—clear entrepreneurial mindset and bias to action. We hired him. He's now making major contributions to one of the biggest apps in the industry.

He didn't have the pedigree. He had proof of work.

What to do about it:

• Before adding any credential filter to a job description, ask: does this role actually require this? Challenge "3 years crypto experience" or "top-tier CS degree" as defaults.

• Change your interview questions. Ask "What have you built that nobody asked you to build?" and "Walk me through a time you learned a new technology quickly and shipped with it."

• Build AI fluency evaluation into every role. Ask candidates: "Walk me through how you use AI tools in your current workflow. When do you use them? When do you not?"

• For technical roles, consider assessment environments that let you observe AI collaboration in real time—how candidates prompt, evaluate outputs, and integrate AI with their own judgment.

• Track the correlation between your credential filters and actual performance. Most companies have never done this analysis. When they do, they find the correlation is weaker than they assumed.

The market is bifurcated. It's an employer's market for generalist roles and a candidate's market for specialized AI and crypto talent. Most founders are applying the wrong market's rules to their hiring. Know which market you're competing in—and remember that today's leverage can become tomorrow's attrition if you don't calibrate intensity intelligently.

Crypto is going enterprise. The M&A numbers tell the story: $2.9B, $1.5B, $1.25B, $1.1B. You need both crypto-native builders who understand primitives and innovation AND engineers with distributed systems depth, strong CS fundamentals, and security expertise. The domain knowledge can be taught. The engineering foundation can't be faked.

Proof of work is beating credentials. McKinsey found 87% of companies face skills gaps—and skills-based hiring adoption has jumped from 40% to 60% since 2020. The ability to ship, learn fast, and leverage AI tools matters more than the logos on a resume. Evaluate accordingly.

Your competitors are still hiring for the market they think they're in. The best founders are hiring for the market that's coming.