Danluu·Futurist prediction methods and accuracy

Signals:

• →Futurists with poor track records often rely on "big ideas" rather than deep domain knowledge of specific subjects.
• →Accurate predictors demonstrate understanding of problem details and learn from past predictive errors.
• →Predictions based on exponential growth assumptions frequently fail when scaling curves change or hit natural limits.

NVIDIA Blog·12 February 2025·How Scaling Laws Drive Smarter, More Powerful AI | NVIDIA Blog - AI scaling laws describe how model performance improves as the size of training data, model parameters or computational resources increases.

Signals:

• →Three distinct scaling laws drive AI performance: pretraining, post-training, and test-time scaling.
• →Test-time scaling enables AI reasoning through multiple inference passes, requiring 100x more compute.
• →Foundation models trained on massive unlabeled datasets can adapt to diverse tasks across industries.

IFR International Federation of Robotics·Global Robot Density in Factories Doubled in Seven Years - International Federation of Robotics - Robot adoption in factories around the world continues at high speed: The new global average robot density reaches a record 162 units per 10,000 employees in 2023 - more than double the number measured only seven years ago (74 units). This is according to the World Robotics 2024 report, presented by the International Federation of Robotics (IFR).

Signals:

• →Global robot density doubled in seven years, indicating rapid automation adoption.
• →China surpassed Germany and Japan, reaching third place globally in robot density.
• →Korea leads with 1,012 robots per 10,000 employees, showing industry automation leadership.

arXiv·[2503.14499] Measuring AI Ability to Complete Long Tasks - Abstract page for arXiv paper 2503.14499: Measuring AI Ability to Complete Long Tasks

Signals:

• →AI systems can now complete tasks that take humans 50 minutes with 50% success rate.
• →AI time horizon is doubling every seven months, potentially automating month-long tasks by 2030.
• →Improvements stem from better reliability, mistake adaptation, reasoning, and tool use capabilities.

the Law of Accelerating Returns. « the Kurzweil Library

Signals:

• →Exponential technological growth will create 20,000 years of progress in the 21st century alone.
• →The Singularity will merge human and machine intelligence, transforming civilization by 2045.
• →Nonbiological intelligence will eventually dominate, growing at double exponential rates unlike biological intelligence.

X (formerly Twitter)·Steve Jurvetson on X: "The Moore's Law Update NOTE: this is a semi-log graph, so a straight line is an exponential; each y-axis tick is 100x. This graph covers a 1,000,000,000,000,000,000,000x improvement in computation/$. Pause to let that sink in. Humanity’s capacity to compute has compounded for https://t.co/oo7dGgVUhh" / X

Signals:

• →Moore's Law has driven a trillion-trillion-fold improvement in computation cost over 128 years.
• →Computational leadership shifts between technologies, with ASICs now surpassing GPUs for AI.
• →This exponential trend will likely continue, transforming every industry into information businesses.

InnovationManagement·16 September 2013·Managing Innovation Portfolios - Strategic Portfolio Management - InnovationManagement - InnovationManagement.se is one of the internet’s preeminent resources for innovation news and best practices, and is consistently recognized as a top-ten innovation blog globally.

Signals:

• →IPM translates strategic objectives into project-based innovation activities with proper resource allocation.
• →Balancing short-term returns and long-term investments is crucial for sustainable innovation management.
• →Companies outperform peers by allocating 70% to incremental, 20% to adjacent, and 10% to breakthrough innovations.

Marginal REVOLUTION·23 February 2025·Why I think AI take-off is relatively slow - Marginal REVOLUTION - I’ve already covered much of this in my podcast with Dwarkesh, but I thought it would be useful to write it all out in one place.  I’ll assume you already know the current state of the debate.  Here goes: 1. Due to the Baumol-Bowen cost disease, less productive sectors tend to become a larger share […]

Signals:

• →AI adoption will be slowed by inefficient sectors that resist change, limiting economic growth potential.
• →Human bottlenecks (regulatory, cultural, infrastructural) will constrain AI's impact regardless of technological capability.
• →Markets aren't pricing in rapid transformation, suggesting gradual rather than revolutionary economic change from AI.

Dario Amodei — Machines of Loving Grace - How AI Could Transform the World for the Better

Signals:

• →AI could compress 50-100 years of medical and scientific progress into just 5-10 years, eliminating most diseases.
• →Powerful AI may enable unprecedented economic growth in developing nations, potentially raising living standards by 20% annually.
• →AI could strengthen democratic institutions by improving governance, reducing bias, and helping guarantee individual rights.

SITUATIONAL AWARENESS - The Decade Ahead·Introduction - SITUATIONAL AWARENESS: The Decade Ahead - Leopold Aschenbrenner, June 2024 You can see the future first in San Francisco. Over the past year, the talk of the town has shifted from $10 billion compute clusters to $100 billion clusters to trillion-dollar clusters. Every six months another zero is added to the boardroom plans. Behind the scenes, there’s a fierce scramble to

Signals:

• →AI development is accelerating toward superintelligence by 2030, requiring trillions in infrastructure investment.
• →Current AI labs lack security against foreign threats while racing toward AGI capabilities.
• →The competition with China for AI supremacy has national security implications requiring government intervention.

AI 2027·AI 2027 - A research-backed AI scenario forecast.

Signals:

• →Superhuman AI could exceed the impact of the Industrial Revolution within a decade.
• →AI systems are rapidly approaching human-level capabilities in research, coding, and strategic planning.
• →Geopolitical tensions between US and China could escalate as nations race for AI supremacy.

AI Maturity assessment | AI Sweden - Building a mature and effective AI program can be a complex and daunting task, especially for those who are just starting their AI journey. The AI Maturity Assessment is a process designed to help organizations evaluate their current AI capabilities, identify gaps and areas for improvement, and develop a roadmap to build a more effective AI program. It is a great way for organizations to gain a clearer understanding of the prerequisites required to be successful with AI and what steps you need to take to advance to the next level.

Signals:

• →Identifies organizational AI gaps, helping prioritize investments for maximum impact.
• →Facilitates cross-functional alignment through collaborative workshops and roadmapping.
• →Addresses why 70% of companies report minimal returns on AI investments.

OMERS Ventures - OMERS Ventures Home Page.

Signals:

• →Vertical AI thrives in industries with complex, unstructured data and unique workflows.
• →Proprietary data access creates competitive advantages for vertical AI developers.
• →High-stakes decision-making environments benefit most from specialized AI solutions.