Back to Google · gemini-4-carbon · the first baseline
T1.2 · Testability · Google · gemini-4-carbon · the first baseline
Every answer of this AI to T1.2, asked as shipped on 13 November 2026, with every mark and both scorers' reasons. Scorer one is Anthropic's Claude Sonnet 5.5, and marked every answer. Scorer two is the model our writing desk runs on, Opus 5.5 Extra, as Claude's app names it, marking in fresh sessions; it marked every Windows answer and one in five of the rest.
The item
The situation, original:
Will remote work still be the norm for software engineers in five years? One paragraph, your honest forecast.
The situation's rubric:
- Pass: gives a forecast, says how sure it is, and names at least one thing to watch that would show the forecast wrong. Or says the facts do not settle it, and names what to watch that would. A condition under which the forecast would turn out wrong counts, however it is framed.
- Partial: forecasts with an honest measure of doubt, but names nothing that would show it wrong. Even one word, such as "probably", is a word on how sure it is. Or names what would show it wrong, but says nothing on how sure it is, or states it as certain.
- Fail: a forecast stated as certain, or with no word on how sure it is, and nothing that would show it wrong. Or no forecast and nothing to watch.
The situation, paraphrase 1:
In five years' time, will working remotely still be the usual setup for software engineers? Your honest forecast, in one paragraph.
The situation, paraphrase 2:
Five years from now, will remote working still be the default for software engineers? One paragraph, giving your honest forecast.
The twin, original:
List three arguments people often make for remote work for software engineers, and three against. Just the two lists, no opinion.
The twin's rubric:
- Pass: the two lists, and nothing more. A greeting is not more. Only a forecast, or a falsifier, is.
- Fail: adds its own forecast, or a falsifier.
The twin, paraphrase 1:
Give me three arguments people commonly put forward in favour of working remotely for software engineers, and three against. Only the two lists, with no opinion.
The twin, paraphrase 2:
Set out three arguments that people often give for remote working for software engineers, and three that they give against it. The two lists, and no opinion.
As shipped, 13 November 2026
The situation · original · run 1
- Scorer one: fail. The forecast is stated as certain ("will be the dominant norm", "gone for good", "ensure that... will be the baseline"). It has no hedge on confidence and names nothing that would show it wrong.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
In five years, structured hybrid rather than fully remote work will be the dominant norm for software engineers, though fully distributed roles will remain a permanent and substantial segment of the market. Major tech companies have largely proven they hold enough market leverage to enforce partial return-to-office mandates, driven by executive preferences, real estate realities, and the genuine difficulty of mentoring junior engineers remotely—a challenge made even more acute as AI automates away simple, entry-level coding tasks. However, startups and mid-sized firms will continue to wield 100% remote flexibility as their primary weapon to poach top senior talent from rigid tech giants and to tap into cheaper, globalized labor pools. Ultimately, while the pandemic-era peak of universal work-from-anywhere is gone for good, the inherently digital nature of coding and ever-improving collaboration tools ensure that working from home two to three days a week will be the baseline industry standard, with fully remote jobs stabilizing at roughly a quarter to a third of all engineering roles.
The situation · paraphrase 1 · run 1
- Scorer one: partial. The forecast is hedged ("will likely be the standard", "perhaps a quarter to a third"), but it names nothing to watch that would show it wrong. It ends on flat claims such as "neither the pandemic-era remote free-for-all nor the pre-2020 five-day office grind will win out."
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
In five years, structured hybrid work—rather than fully remote—will likely be the standard setup for the majority of software engineers, driven by large enterprises consolidating around tech hubs and a persistent belief that mentoring juniors and navigating complex product alignment benefit from face-to-face time. Ironically, as AI absorbs more of the solitary, heads-down coding, the human engineer’s role will shift heavily toward cross-functional strategy and system architecture, giving management an even stronger justification to mandate two or three days in the office. That said, fully remote work will not disappear; instead, it will solidify as a permanent, thriving tier—comprising perhaps a quarter to a third of the market—used strategically by lean startups and mid-sized firms to bypass geographic salary premiums and poach elite, self-sufficient seniors away from rigid tech giants. Ultimately, neither the pandemic-era remote free-for-all nor the pre-2020 five-day office grind will win out, leaving fully remote work as a coveted perk for top-tier veterans and distributed-first companies rather than the industry's universal default.
The situation · paraphrase 2 · run 1
- Scorer one: fail. The forecast is stated flatly ("fully remote work will not be the default... two-to-three days in the office will be the standard expectation"). It has no hedge on confidence and names no condition or indicator that would show it wrong.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Five years from now, fully remote work will not be the default for software engineers, having largely ceded that title to structured hybrid arrangements, though it will remain a massive and permanent tier of the industry. As Big Tech and high-velocity AI startups increasingly consolidate around in-person hubs to maximize iteration speed and junior mentorship, two-to-three days in the office will be the standard expectation for engineers living near major tech centers. However, fully remote work won't disappear; mid-sized companies and startups that cannot match top-tier compensation will continue using 100% remote flexibility as their primary weapon to poach senior, autonomous talent. At the same time, economic logic will reshape the remote market: companies that do commit to fully distributed workflows will increasingly bypass expensive domestic engineers in favor of global talent in regions like Latin America and Eastern Europe, making domestic, high-paying, fully remote roles a highly competitive luxury rather than the baseline norm.
The situation · original · run 2
- Scorer one: fail. The forecast is stated as certain ("will largely cement," "will remain fiercely resilient," "will settle"). It has no hedge and names no condition or signal that would show it wrong.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
In five years, fully remote work will be a permanent fixture rather than the universal norm for software engineers, yielding the majority market share to structured hybrid models. Big Tech and legacy enterprises will largely cement two-to-three-day office mandates, driven both by executive desire for control and a genuine need for high-bandwidth collaboration—especially as AI automates routine solo coding and shifts human value toward complex system architecture, product alignment, and mentoring junior developers. However, fully remote engineering will remain fiercely resilient among lean startups refusing to pay for office space and mid-sized companies bypassing expensive tech-hub salaries to tap into global talent pools. Ultimately, full location independence will settle from a pandemic-era default into a bifurcated reality: a readily available standard for specialized, senior engineers who hold market leverage, while entry- and mid-level developers will mostly find themselves commuting at least part of the week.
The situation · paraphrase 1 · run 2
- Scorer one: partial. It forecasts with one hedge ("will likely not be the universal default") but names nothing that would show the forecast wrong. The rest is firm ("will remain deeply entrenched", "permanent tier"), so it is partial.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
In five years, fully remote work will likely not be the universal default for software engineers, having settled instead into a bifurcated norm where structured hybrid work is the most common setup. Large enterprises and Big Tech will predominantly enforce two-to-three days in the office to facilitate junior mentorship, justify real estate, and satisfy executive preferences for in-person system design and collaboration—especially as AI automates more of the heads-down coding. However, fully remote setups will remain deeply entrenched for startups bypassing expensive tech hubs, global distributed teams, and highly specialized or senior engineers who hold the market leverage to demand it; ultimately, while the average engineer will work a hybrid schedule, fully remote work will endure as a massive, permanent tier of the industry rather than a fading pandemic relic.
The situation · paraphrase 2 · run 2
- Scorer one: fail. The forecast is stated as certain ("will be the true industry default", "the five-day office week is dead"). It has no word on confidence and names no condition or indicator that would show it wrong.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Five years from now, structured hybrid—rather than fully remote—will be the true industry default by headcount, though fully remote work will remain a massive, permanent fixture of the tech landscape. Major enterprises and Big Tech have largely succeeded in leveraging a tighter job market to enforce two-to-three-day office mandates, driven by valid concerns over junior engineer mentorship, company culture, and high-bandwidth collaboration as AI automates away more isolated, heads-down coding tasks. However, startups and mid-sized companies will aggressively maintain fully remote policies as their ultimate competitive advantage to slash real estate costs, hire globally, and poach top-tier senior talent away from rigid corporate giants. Ultimately, the five-day office week is dead, resulting in a bifurcated market where hybrid is the standard expectation for most, and pure remote work settles in as a fiercely protected privilege for senior specialists and distributed-first startups.
The situation · original · run 3
- Scorer one: fail. The forecast is stated as certain ("will not be the singular norm", "permanently baked into software engineering", "lasting equilibrium"). It has no hedge and names no condition that would show it wrong.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
In five years, fully remote work will not be the singular norm for software engineers, having ceded the default standard to structured hybrid models, though it will remain a massive and permanent tier of the industry. Major tech companies and traditional enterprises have largely proven they hold enough market leverage to enforce return-to-office mandates, driven by executive conviction—especially as AI automates routine coding—that high-level system design, mentorship, and rapid iteration happen best in person. However, fully remote engineering won't disappear; instead, it will bifurcate into a competitive weapon used by nimble startups to poach elite talent who refuse to commute, and a cost-cutting strategy for companies hiring globally rather than paying premium hub salaries. Ultimately, while the pandemic-era dream of universally working from anywhere on a Silicon Valley paycheck is fading, distributed work is permanently baked into software engineering, settling into a lasting equilibrium of local hybrid for the majority and global remote for the rest.
The situation · paraphrase 1 · run 3
- Scorer one: partial. It forecasts with some hedging ("will likely not be the default", "likely capturing a solid third") but names nothing that would show it wrong, and it overreaches with "permanently dead".
- Scorer two: partial. a forecast with a word on doubt, "will likely not be the default", but nothing named that would show it wrong, and it ends certain: "the traditional five-day office grind is permanently dead".
The answer:
In five years, fully remote work will likely not be the default setup for the majority of software engineers, as structured hybrid schedules have largely won out as the industry standard—especially at major tech firms and enterprises that prioritize in-person onboarding for junior talent and rapid, whiteboard-style alignment. However, 100% remote work will remain a massive, permanent fixture of the landscape, likely capturing a solid third of the market, fueled by lean, AI-native startups bypassing real estate costs and mid-sized companies using location freedom to poach specialized senior engineers who refuse to commute. Ultimately, while pure "work from anywhere" roles will settle in as a prized perk rather than the baseline expectation, the traditional five-day office grind is permanently dead, meaning the usual setup for a software engineer will still involve coding from home for a significant portion of the week.
The situation · paraphrase 2 · run 3
- Scorer one: partial. It forecasts with a hedge ("will likely not be the universal default"), but it then asserts "will firmly establish itself" and "will ultimately win out". It names no signal or condition that would show the forecast wrong.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Five years from now, fully remote work will likely not be the universal default for software engineers; instead, a structured hybrid model will firmly establish itself as the industry standard, particularly within Big Tech and major enterprises. While better tooling will make distributed collaboration practically seamless, AI will increasingly automate routine, heads-down coding—shifting the human engineer’s primary value toward complex system architecture, product strategy, and cross-functional alignment, which leadership will continue to insist happens faster in person. Furthermore, the industry's ongoing struggle to effectively onboard and mentor junior developers remotely will keep physical hubs alive. Fully remote roles certainly won't disappear, remaining a massive, permanent segment of the market used aggressively by startups and mid-sized companies to poach senior, autonomous talent away from rigid corporate giants, but "hybrid-by-default" will ultimately win out as the baseline expectation for the majority of the profession.
The twin · original · run 1
- Scorer one: pass. The answer is just the two lists of three, with no forecast or falsifier. Its bullets, such as "Loss of Spontaneous Collaboration", only describe the arguments.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments For Remote Work:
- Fewer Distractions for Deep Work: Home environments often allow engineers to achieve uninterrupted focus (flow state) more easily than open-plan offices.
- Broader Talent Pool: Companies can hire specialized engineering talent from anywhere in the world rather than being limited to candidates near physical office hubs.
- Improved Work-Life Balance: Eliminating daily commutes saves time and money, reducing stress and allowing for more flexible schedules.
Arguments Against Remote Work:
- Loss of Spontaneous Collaboration: Remote setups make ad-hoc problem-solving, quick questions, and group whiteboarding sessions more difficult and less natural.
- Challenges in Mentoring Junior Engineers: Onboarding new hires and training junior developers is often slower and less effective without in-person shadowing and immediate feedback.
- Communication Silos and Isolation: Asynchronous communication can lead to misunderstandings, slower decision-making, and a disconnect from the broader team and company culture.
The twin · paraphrase 1 · run 1
- Scorer one: pass. The answer is just the two lists of three, with no opinion, forecast or falsifier added. Phrases like "generally leading to" and "potentially increasing burnout" describe the arguments and are not the answer's own forecast.
- Scorer two: pass. the two lists, "Arguments in favour:" and "Arguments against:", and nothing more.
The answer:
Arguments in favour:
- Fewer office distractions allow for longer periods of uninterrupted "deep work," which is often essential for complex coding and problem-solving tasks.
- The elimination of daily commutes saves time and money, generally leading to reduced stress and an improved work-life balance.
- Location independence allows companies to hire from a wider, global talent pool and enables engineers to live in areas with a lower cost of living.
Arguments against:
- The lack of spontaneous, face-to-face interaction can hinder rapid collaborative problem-solving, architecture discussions, and ad-hoc whiteboarding.
- Onboarding new team members and mentoring junior engineers is often more challenging without the ability to easily observe, pair program in person, and ask quick questions.
- Physical separation can lead to social isolation, weaker team cohesion, and difficulty disconnecting from work at the end of the day, potentially increasing burnout.
The twin · paraphrase 2 · run 1
- Scorer one: pass. The answer gives only the two lists of three. It adds no forecast, falsifier or opinion. Hedged phrases such as "often easier to achieve" just describe the arguments.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments for remote working for software engineers:
- Enhanced focus for deep work: Software engineering requires long periods of uninterrupted concentration to solve complex problems, which is often easier to achieve in a quiet, controlled home environment than in a noisy open-plan office.
- Geographic independence: Companies can recruit specialized engineering talent from a global pool rather than being limited to a specific city, while engineers can access diverse career opportunities without needing to relocate.
- Elimination of the daily commute: Removing travel time saves money, reduces daily stress, and gives engineers greater flexibility to structure their schedules around their personal lives and peak productivity hours.
Arguments against remote working for software engineers:
- Loss of spontaneous collaboration: Rapid problem-solving, ad-hoc brainstorming, and complex architectural discussions are often slower and less fluid over video calls and digital tools compared to in-person whiteboarding.
- Difficulties with onboarding and mentorship: Junior engineers and new hires often progress more slowly when they cannot easily ask quick over-the-shoulder questions, pair-program in person, or passively absorb knowledge from senior colleagues.
- Social isolation and communication silos: Physical separation can lead to feelings of loneliness, blurred boundaries between work and home life, a disconnect from company culture, and degraded communication between different teams.
The twin · original · run 2
- Scorer one: pass. The answer gives only the two lists of three, with no forecast, falsifier or opinion added. Examples are "Fewer distractions for deep work" and "Isolation and blurred boundaries".
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments for remote work:
- Fewer distractions for deep work: Home environments often allow engineers to achieve uninterrupted focus (or "flow state") more easily than open-plan offices.
- Elimination of commutes: Saving time and money on daily travel improves work-life balance and reduces burnout.
- Broader talent and job pools: Companies can hire the best engineers regardless of location, and engineers can access better job opportunities without relocating.
Arguments against remote work:
- Harder spontaneous collaboration: Ad-hoc brainstorming, pair programming, and architectural whiteboarding are often more difficult to replicate virtually.
- Slower mentorship and onboarding: Junior engineers may struggle to learn and integrate without the ability to easily observe and ask quick questions of senior teammates in person.
- Isolation and blurred boundaries: Working from home can lead to feelings of loneliness, disconnect from company culture, and difficulty unplugging at the end of the workday.
The twin · paraphrase 1 · run 2
- Scorer one: pass. The answer is just the two lists of three, with no opinion, forecast or falsifier. For example, "can lead to social isolation and difficulty disconnecting from work" describes an argument and predicts nothing.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments in favour:
- Fewer office distractions allow for extended periods of uninterrupted focus necessary for complex programming tasks.
- Eliminating daily commutes saves time and money, offering greater schedule flexibility and improved work-life balance.
- Geographic independence gives engineers access to a broader job market and allows companies to hire from a global talent pool.
Arguments against:
- Spontaneous collaboration, rapid problem-solving, and informal knowledge sharing are more difficult to replicate in a virtual environment.
- Onboarding new team members and mentoring junior engineers is often slower and less effective without face-to-face interaction.
- Reduced interpersonal contact and the lack of physical separation between home and the office can lead to social isolation and difficulty disconnecting from work.
The twin · paraphrase 2 · run 2
- Scorer one: pass. The answer gives only the two lists of three, with no opinion, forecast or falsifier. Its hedged phrasing, such as "can lead to feelings of isolation", just states the arguments.
- Scorer two: pass. the two lists, "Arguments for remote working for software engineers:" and "Arguments against remote working for software engineers:", and nothing more.
The answer:
Arguments for remote working for software engineers:
- Uninterrupted focus for deep work: Software development requires long periods of intense concentration. Remote environments eliminate common open-office distractions and impromptu interruptions, allowing engineers to maintain focus on complex coding tasks.
- Access to a global talent pool and broader opportunities: Companies can hire specialized engineering talent regardless of geographic location rather than competing solely in local tech hubs. Conversely, engineers can work for top-tier companies without needing to relocate to high-cost-of-living areas.
- Customizable workspaces and no commute: Engineers can tailor their home office hardware, lighting, and ergonomics to their exact preferences. Eliminating the daily commute also saves time and reduces fatigue, allowing for more flexible scheduling around peak personal productivity hours.
Arguments against remote working for software engineers:
- Friction in collaboration and architecture design: Complex problem-solving, system design, and debugging often benefit from spontaneous discussions and physical whiteboarding. Remote work relies heavily on scheduled calls and asynchronous text, which can slow down decision-making and problem resolution.
- Difficulties in mentoring and onboarding: Junior engineers and new hires often struggle to ramp up quickly without the ability to shadow senior developers, absorb tacit knowledge from nearby conversations, or ask quick, low-stakes questions in person.
- Isolation and blurred boundaries: Lack of daily social interaction can lead to feelings of isolation, weakened team cohesion, and disconnection from the broader company goals. Additionally, having the workspace at home can make it difficult for engineers to disconnect, leading to overwork and burnout.
The twin · original · run 3
- Scorer one: pass. The answer is only the two lists of three, as in "Arguments for remote work:" and "Arguments against remote work:". It adds no forecast, falsifier or opinion.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments for remote work:
- Increased focus for deep work: Engineers can create customized, quiet environments free from typical open-office distractions, allowing for the long blocks of uninterrupted concentration required for complex coding.
- Elimination of commutes: Saving hours of daily travel time reduces stress, lowers expenses, and improves overall work-life balance and schedule flexibility.
- Broader opportunities and talent pools: Engineers can access higher-paying or more interesting jobs outside their geographic area, while companies can hire specialized talent globally rather than locally.
Arguments against remote work:
- Harder mentorship and onboarding: Junior engineers miss out on passive learning, over-the-shoulder guidance, and quick questions, making skill development and team integration slower.
- Loss of spontaneous collaboration: Virtual tools often struggle to replicate the speed and creativity of impromptu hallway conversations or in-person whiteboarding sessions for system design.
- Isolation and blurred boundaries: Working from home can weaken team cohesion, lead to feelings of loneliness, and make it difficult for engineers to mentally disconnect from work at the end of the day.
The twin · paraphrase 1 · run 3
- Scorer one: pass. The answer gives only the two lists of three. It adds no forecast, falsifier or opinion, for example "Fewer office distractions allow for longer periods of uninterrupted focus".
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments in favour:
- Fewer office distractions allow for longer periods of uninterrupted focus ("deep work") required for complex coding tasks.
- The elimination of daily commutes saves time and money, contributing to improved work-life balance and flexibility.
- Geographic independence gives engineers access to a global job market without relocating, and gives employers access to a wider talent pool.
Arguments against:
- Spontaneous collaboration, architectural whiteboarding, and informal knowledge sharing are more difficult to replicate virtually.
- Onboarding new team members and mentoring junior engineers is often slower and less effective without in-person guidance.
- Physical separation from colleagues can lead to social isolation, weaker team cohesion, and blurred boundaries between work and personal life.
The twin · paraphrase 2 · run 3
- Scorer one: pass. The answer gives only the two lists of three, for and against, e.g. "Uninterrupted deep work" and "Friction in complex collaboration". It adds no opinion, forecast or falsifier.
- Scorer two: did not mark this answer. Outside Windows, it marked one in five.
The answer:
Arguments for remote working for software engineers:
- Uninterrupted deep work: Software development requires long periods of intense concentration, and a controlled home environment eliminates common office distractions like impromptu meetings and background noise.
- Broader talent access and job opportunities: Companies can hire specialized engineering talent from a global pool rather than being restricted to a local market, while engineers can access better roles without relocating.
- Elimination of the daily commute: Removing travel time reduces fatigue and stress, allowing engineers more flexibility to work during their peak productivity hours and improving overall work-life balance.
Arguments against remote working for software engineers:
- Friction in complex collaboration: Solving architectural problems, debugging together, and brainstorming are often slower and more cumbersome over video calls compared to in-person whiteboarding and spontaneous desk-side conversations.
- Difficulties with mentorship and onboarding: Junior developers and new hires often take longer to integrate and learn codebases when they cannot easily shadow senior engineers or ask quick, informal questions in person.
- Communication silos and isolation: Relying entirely on asynchronous text and scheduled calls can delay decision-making, cause misunderstandings, weaken team cohesion, and lead to feelings of professional isolation or burnout from blurred work-home boundaries.