BAE1. argmax_x (u|x, theta's atomness) industry effect

[hyunjimoon]

table.ENVIRONMENT1

Aspect	DIGITAL	PHYSICAL	BIOTECH	SEMICONDUCTOR	TRANSPORTATION/MOBILITY
Development	Software, apps, and/or online services	Develop and set up products and processes, including sourcing raw materials, developing supply chains, and production lines.	Focused on research & development in labs, clinical trials for drug discovery, genetic engineering.	Design and development of semiconductor devices, prototyping, and engineering for manufacturing.	Design, development, and integration of mobility solutions (vehicles, infrastructure, technology systems).
Production	Upfront cost of development, servers, and other digital infrastructure	Continuous management of manufacturing & supply chains	Small scale for clinical trials, can scale up with contract manufacturing organizations.	High initial setup costs for fabs, ongoing costs for materials, and precision manufacturing processes.	Initial high costs for infrastructure and vehicle production, ongoing maintenance, and operational costs.
Cost Structure/ Scalability	~Zero marginal costs for more copies; May require increased marketing reach	Unit production costs can be significant and increasing scale may require added capital investment	Lower marginal costs for drug production once approved; scalability dependent on regulatory approval.	Significant investment in fabrication plants; economies of scale important for cost reduction.	High initial investment in infrastructure, vehicles, technology; scalability through network expansion and efficiency improvements.
Distribution	Online via app stores, websites, and/or cloud-based services	Transportation to retailers or directly to consumers; requires logistics, inventory management, returns, etc.	Distribution through healthcare providers, pharmacies; regulatory compliance for market access.	Distribution through technology product manufacturers, direct sales to electronics companies.	Through mobility services (public transport, ridesharing), vehicle sales, and infrastructure deployment.
Updates and Improvements:	Can be continuous, remote, and as visible or invisible as desired without the user even noticing. Allows for rapid iteration based on user feedback and data analytics.	Improvements or modifications require creating a new version of the product and potentially recalling/repairing the old one.	Updates through additional research, new product versions; continuous improvement in treatment efficacy.	Product iterations require new manufacturing cycles; updates often coincide with new product releases.	Continuous service improvement, technology updates, infrastructure upgrades; may involve regulatory approvals.
Data Collection and Analytics	Data collection and analytics can be engineered into the products and infrastructure to a significant degree	Unless the product has a digital/connected component, data collection may be challenging, especially when the product was not sold directly be the manufacturer.	Increasing use of digital health technologies for data collection in clinical trials and patient monitoring.	Embedded sensors and IoT integration enable data collection for some semiconductor products.	Data from mobility services and smart infrastructure for optimization, service improvement, and user experience enhancement.
End of Life	Discontinue support or updates, shut down servers or online services. Data migration and user communication are key considerations. Some labor redeployment.	Discontinue production and factories, and dispose of unsold inventory. Environmental considerations, such as recycling or disposal of the product, are also important. Significant layoffs possible.	Discontinuation of older drugs, biologics; focus shifts to newer treatments. Environmental impact of disposal considered.	End of product lifecycle management, recycling of electronic components; significant environmental considerations.	Decommissioning of vehicles and infrastructure, environmental impact mitigation, recycling, and repurposing efforts.
Entry Barriers	Competitors may be able to enter more easily. Including pirates.	Manufacturing and supply chain capabilities may create higher barriers to entry	High due to regulatory hurdles, IP protections, and the need for specialized knowledge.	Very high due to the cost of fabrication plants, specialized equipment, and technical expertise.	High due to significant capital requirements, regulatory approvals, and need for technological expertise.
Cell Theory	Platforms enabling individual developers to focus on niche digital solutions.	Startups focusing on developing specific hardware technologies.	Individual biotech firms focusing on niche healthcare solutions.	Startups developing specific semiconductor technologies.	Development and integration of technologies for enhancing mobility, including electric vehicles, autonomous systems, and smart infrastructure. Collaborative ventures aiming at creating innovative transportation solutions.
Ecosystem	Digital marketplaces and developer communities supporting collaboration.	Industrial ecosystems involving suppliers, manufacturers, and distributors.	Collaboration with healthcare systems, regulatory bodies, and research.	Interactions with suppliers, manufacturers, and tech companies.	A dynamic ecosystem including transportation providers, technology companies, regulatory bodies, and consumers. Encourages public-private partnerships.
Evolution	Rapid iteration and innovation in software based on user feedback.	Gradual technological advancements through R&D and market feedback.	Innovation in response to healthcare challenges and regulations.	Technological advancements and market adaptation.	Continuous innovation driven by technological advancements, regulatory changes, and user demands. Focus on sustainability, safety, and accessibility.
Phylogeny	Tracing the development and influence of major software and internet technologies.	Tracing the history and evolution of manufacturing processes and materials.	Tracing the development of critical medical technologies.	Evolution of chip manufacturing techniques.	Tracing the evolution of transportation technologies and systems, from early mobility solutions to modern autonomous and electric vehicles.
Properties of Life	Digital products growing through user adoption, evolving with market needs.	Physical products growing through market penetration, evolving with manufacturing innovations.	Growth through research progress, reproduction via spin-offs, response to healthcare trends.	Growth through technological innovation, reproduction through patents, adaptation to tech trends.	Growth through technological adoption and infrastructure development, reproduction through new service models, response to societal mobility needs.

This additional column on Transportation/Mobility reflects the integration of mobility solutions into the framework of development, production, distribution, and other key aspects as derived from the course materials on Mobility Ventures.

[hyunjimoon]

decisions are based on efficacy, safey, market size, competitor

The first plot (top) includes:
Manufacturing Cost Evolution ($/Patient) in red
Total Patients for Clinical Test (Patient) in blue dashed line
Total Patients (Patient) in blue solid line
The second plot (bottom) includes:
Manufacturing Costs for Clinical Test ($) in green solid line
Drug Manufacturing Cost ($) in green dashed line

decision tree allows a biotech startup to choose theta (investment, exit) based on information below (patient#, live births, incidence, expected peak penetration, reimbursement, penetration, market,

Manufacturing costs:=G64(total patients)G29(cost per patient)
Drug manufacturing costs for clincial test:= patient number(sma1 + sma2) cost per patient
once test is finished for both types of patients (2021), then it starts applying to more pateinets(2022-2031).

[hyunjimoon]

is big bump from fixed factory cost?

[hyunjimoon]

Variable	Pure Digital	PhD in Academia doing empirical research	Biotech Startup	Semiconductor Startup	Pure Physical
Candidates	Digital product concepts, beta versions	Prototype ideas	Potential therapeutic compounds, early-stage research	Semiconductor device concepts, early designs	Product designs, prototypes
Products	Released software, apps, online services	Paper combining hypothesis, measured phenomena	Developed drugs, clinical trial compounds	Fabricated semiconductor devices	Finished goods, manufactured items
Used Product	Software and services in active use by consumers	Published and read paper	Medications being taken by patients, biotech products in use	Semiconductor devices integrated into products and used by consumers	Products being used by consumers
Pairing Time	Time to match software with market/user needs	Prototyping time	Time to align research with therapeutic targets	Time to develop and design semiconductors	Time to design and prototype
Production Time	Time spent in software development and testing	Experiment time	Time for conducting experiments, trials, and production of compounds	Time for prototyping, manufacturing, and testing semiconductor devices	Time spent in manufacturing and quality control
Delivering Time	Time for software deployment and distribution	Research diffusion time	Time for drug approval, production scaling, and delivery to providers	Time from manufacturing to delivery to tech companies or integration into products	Time for logistics, from factory to consumer
Life Time	Duration of software utility before becoming obsolete or updated	Paper life time	Duration from drug release to obsolescence due to new therapies or expiration of patent	Lifecycle of semiconductor device before becoming obsolete due to new technology	Duration of product usability before wear out or obsolescence

This table is now titled "Supply Chain in Industries with Different Hardwareness" and reflects the different stages and timeframes associated with product development and lifecycle in various industries, from purely digital to purely physical, including the academic process of empirical research.

[hyunjimoon]

Variable	Pure Digital	PhD in Academia doing empirical research	Biotech Startup	Semiconductor Startup	Pure Physical
Sales (Usefulness)	Number of users, subscriptions, downloads, user engagement metrics	Citations, academic impact, paper downloads, conference presentations	Drug sales, usage by patients, citation in treatment guidelines	Number of chips sold, use in prominent tech products	Sales volume, customer satisfaction, repeat purchases
Asset Allocation	Allocation of capital to software development, digital infrastructure, marketing	Research resources (time, talent, effort, money), allocation of funding towards specific research areas	Allocation of funding to R&D, clinical trials, and market preparation	Investment in R&D, manufacturing equipment, and materials	Allocation of funds to manufacturing, logistics, and product development
Demand Develop Asset	Investment in market research, user experience design, analytics	Resources used to collect and address needs in academia (e.g., identifying emerging social issues, gathering data)	Resources allocated to understanding patient needs, market research for new treatments	Resources dedicated to market analysis for new semiconductor applications	Market research to identify consumer needs and trends
Supply Develop Asset	Resources dedicated to technological innovation, acquiring digital talent	Resources used to develop supply in academia (e.g., building expertise or methodology for research, expanding collaborator networks)	Investment in technology, expertise acquisition, and partnerships for drug development	Investment in developing technological expertise, fabrication methods, and supplier networks	Investment in improving production capabilities, supplier relationships, and distribution networks
Reinvest Reinforcing Loop	Reinvestment from profits into product development, market expansion	Reinvesting in the research area that gains traction through citations, which can lead to increased funding and resources, thereby enhancing research capabilities and capacities	Profits reinvested into R&D for new treatments, enhancing production capabilities	Reinvestment of earnings into R&D for more advanced technologies, improving manufacturing capabilities	Reinvestment from sales into product improvement, manufacturing efficiency, and market expansion
Unaddressed Need	User needs or market segments not yet met by current digital offerings	Unmet research needs or areas where existing theories and methodologies do not suffice	Patient needs unmet by current treatments or drugs	Technological needs in the market not yet fulfilled by existing semiconductor technologies	Consumer needs not yet met by available products
Unused Solution	Digital innovations that haven't found a market fit or user base	Research findings or methodologies that haven't been applied or widely recognized in academia	Biotech solutions developed but not yet brought to market due to various constraints (e.g., regulatory approval)	Semiconductor technologies developed but not yet integrated into products due to market readiness or compatibility issues	Products developed but not yet utilized in the market or obsolete due to innovation elsewhere

This table attempts to capture the flow from development to market impact within different industries, considering how resources are allocated and reinvested to grow capabilities and address both the needs and solutions available within each sector.

[hyunjimoon]

I explained in startup_action|industry_hardwareness.pdf, step by step, how startup create value (operation, strategy), captures value. The last page explains how ecosystem should be designed to support efficient startup operation, which depends on hardwareness of industry.

Limiting the comparison to academia, even among this relatively pure software industry where knowledge is produced and delivered , further spectrum exists between social science (software) and physics (hardware). I'm tracing why measurement tool improved in

Matt Cronin suggests detailed modeling in theory crisis in management research and his paper (with stock flow diagram with novelty component added) details of "producing " flow variable and testing causal loop.

Another interesting Matt's comment is "there are a lot of things that we don't understand that will help us with problems later" which reminds me of exapation, logged in #192

[hyunjimoon]

neural network one layer = unit theory.

one at a time (row generation); increasing dimensionality (can't eliminate )
input and parameterization of model; Double descent; generalizability
high dimensional (causal density) input makes modeling hard (theorize)
high dimensional parameter space makes model estimation easy
variables involved in determining outcome
connections between them are dense

low dimensional input, low parameter space
fewer variables (hard to explain data)

[hyunjimoon]

Based on three example analysis in the table below, gpt drew comparison plot on time and money investment on AGCT actions of nail, scale phase (need to verify with experts):

Inferred comparison of two industries are based on generic definition of lifecycle phase and action taxonomy.

1.NSS startup lifecycle: Nail phase focuses on establishing a deep understanding of the target market and refining the value proposition to ensure the business model addresses market needs effectively, aiming for a validated business model canvas. The Scale phase emphasizes expanding operational capabilities, product offerings, and market reach, aiming for sustainable growth through operational excellence and market penetration. The Sail phase concentrates on leveraging stability and infrastructure to explore new growth opportunities, emphasizing agility and continuous innovation in response to changing market condition

2.ATGC action taxonomy: classified across four areas Atom supply/architecture (A), Technology/organization (T), Goal (G), and Customer (C), which are further evaluated by operational feasibility, technology feasibility, employee desirability, customer desirability

Nail Stage Differences

Biotech Industry:

A actions: High resource intensity, both in time and money, due to the need for establishing labs and partnerships. Ratio: 30% Time, 40% Money.
G actions: Moderate, focused on securing funding and defining long-term objectives. Ratio: 10% Time, 10% Money.
C actions: Critical for identifying unmet medical needs, requiring significant market research. Ratio: 25% Time, 20% Money.
T actions: High, as R&D to develop and validate biological technologies is resource-intensive. Ratio: 35% Time, 30% Money.

Semiconductor Industry:

A actions: Moderate, with a focus on forming ecosystems and initial technology development setups. Ratio: 20% Time, 25% Money.
G actions: Moderate, as setting strategic goals and early collaborations take precedence. Ratio: 15% Time, 15% Money.
C actions: Less intensive compared to biotech, with a focus on market gap analysis. Ratio: 15% Time, 10% Money.
T actions: High, due to the emphasis on developing unique semiconductor technologies (e.g., processors, IPUs, silicon photonics). Ratio: 50% Time, 50% Money.

Scale Stage Differences

Biotech Industry:

A actions: Continued high investment in scaling manufacturing capabilities and regulatory compliance. Ratio: 40% Time, 50% Money.
G actions: Moderate, with an emphasis on expanding the product pipeline and entering new markets. Ratio: 10% Time, 10% Money.
C actions: Increased efforts in clinical trials and market validation. Ratio: 25% Time, 20% Money.
T actions: Sustained R&D investment to enhance and broaden technology applications. Ratio: 25% Time, 20% Money.

Semiconductor Industry:

A actions: Significant resources devoted to scaling production and enhancing fabrication technologies. Ratio: 30% Time, 40% Money.
G actions: Moderate, with strategic shifts towards new computing paradigms and applications. Ratio: 10% Time, 10% Money.
C actions: Moderate, expanding into new markets like AI, automotive, and IoT. Ratio: 20% Time, 15% Money.
T actions: High, reflecting continuous innovation in semiconductor technologies and product diversification. Ratio: 40% Time, 35% Money.

Observations:
Biotech startups tend to allocate a larger portion of resources to A and T actions across both stages, driven by the necessity of establishing substantial physical infrastructure and conducting extensive R&D.
Semiconductor startups exhibit a strong emphasis on T actions, indicative of the sector's rapid technological evolution and the critical importance of innovation.
C actions are more emphasized in the biotech industry during the Nail stage, reflecting the need to closely align product development with specific medical needs and regulatory requirements.
A actions in the Scale stage show that semiconductor companies invest heavily in production capabilities and advanced manufacturing technologies to maintain competitiveness.

Company	Nail Phase	Scale Phase	Sail Phase
Moderna	C (2010): Focused on infectious diseases for mRNA applications. T (2010): R&D to overcome mRNA stability and delivery challenges. A (2010): Established foundational infrastructure and partnerships. C (2011): Market research to refine vaccine targets. T (2012): Developed proprietary lipid nanoparticles for delivery.	A (2015): Built manufacturing facilities for scale-up. C (2016): Initiated Phase I trials, validating market need. T (2017): Enhanced mRNA platform for broader vaccine applications. A/C (2018): Secured government and health organization contracts. A/T (2019): Expanded distribution and production capacity in anticipation of COVID-19 vaccine.	G (2020): Pivoted rapidly to develop COVID-19 vaccine. A (2021): Global expansion of manufacturing and distribution. T (2021): Development of mRNA-1010 flu vaccine. C (2022): Adaptation to COVID-19 variants with booster shots. G (2023): Investment in R&D for non-infectious disease applications.
Genentech	C (1976): Targeted critical diseases lacking effective treatments. T (1978): Pioneered recombinant DNA technology for human insulin. A (1978): Setup initial collaborations with academic institutions. C (1980): Conducted foundational research on monoclonal antibodies. T (1982): Developed and commercialized human growth hormone.	A (1985): Expanded production facilities for biopharmaceutical manufacturing. C (1990): Launched innovative drugs, increasing market presence. T (1993): Introduced first antibody-based cancer treatment. A/C (1995): Formed strategic partnerships for global drug distribution. A/T (1999): Increased R&D investments for pipeline diversification.	G (2009): Emphasized oncology and personalized medicine post-Roche acquisition. T (2012): Launched targeted therapies for specific cancer mutations. A (2014): Upgraded manufacturing for biologics production. C (2016): Expanded into immunotherapy with atezolizumab. G (2018): Invested in digital health for R&D innovation.
CRISPR Therapeutics	C (2013): Identified genetic disorders as primary targets. T (2013): Invested in optimizing CRISPR-Cas9 for therapeutic use. A (2014): Established key pharma and biotech partnerships. C (2015): Engaged with patient advocacy groups for disease insights. T (2016): Demonstrated proof of concept for sickle cell and beta thalassemia.	A (2017): Partnered with CMOs for gene therapy production. C (2018): Launched Phase I/II trials for primary target diseases. T (2019): Expanded CRISPR platform capabilities. A/C (2020): Formed strategic distribution partnerships. A/T (2020): Ramped up manufacturing and R&D facilities for scaling.	T/A (2021): Advanced to Phase III trials, showcasing clinical efficacy. C (2021): Entered CAR-T therapy market for cancer treatment. A (2022): Broadened collaborations for new therapeutic areas. G (2023): Focused on global accessibility of gene therapies. T (2024): Continued innovation in CRISPR technology for precision editing.

Company	Nail Phase	Scale Phase	Sail Phase
SiFive, Inc.. .	C (2015): Identified market gap for customizable processor cores. T (2015): Developed RISC-V-based processor IP. A (2016): Formed partnerships within the semiconductor ecosystem. C (2016): Engaged with open-source community for feedback. T (2017): Demonstrated first RISC-V SoC platform.	A (2018): Expanded engineering teams and infrastructure. C (2019): Entered collaborations with semiconductor giants. T (2019): Launched next-gen processor cores. A/C (2020): Developed scalable licensing model. A/T (2020): Enhanced IP portfolio for computing needs.	G (2021): Focused on AI and IoT markets with RISC-V. A (2022): Invested in next-gen fabrication technologies. T (2022): Introduced custom silicon solutions. C (2023): Expanded into automotive and 5G markets. G (2024): Aimed for leadership in sustainable computing.
Graphcore	C (2016): Identified demand for specialized AI hardware. T (2016): Developed Intelligence Processing Unit (IPU). A (2017): Built partnerships with AI research and tech companies. C (2017): Focused on AI developers' requirements. T (2018): Set new standards for AI computation with IPU.	A (2019): Scaled production with semiconductor manufacturers. C (2020): Collaborated with cloud providers for IPU services. T (2020): Launched second-gen IPUs. A/C (2021): Expanded global sales and support teams. A/T (2021): Diversified into hardware and cloud-based AI services.	G (2022): Aimed to standardize AI hardware in research/enterprise. A (2023): Leveraged advanced manufacturing for IPU efficiency. T (2023): Developed software frameworks for easier AI development. C (2024): Entered partnerships in autonomous vehicles and health tech. G (2024): Committed to innovation with a focus on ethical AI.
Ayar Labs	C (2015): Identified bandwidth and power efficiency challenges. T (2015): Developed silicon photonics for optical I/O solutions. A (2016): Co-developed optical I/O with industry partners. C (2017): Aligned product development with market needs. T (2018): Integrated optical I/O with semiconductor platforms.	A (2019): Expanded R&D for optical I/O technology. C (2020): Piloted optical I/O with cloud providers. T (2020): Launched TeraPHY™ chiplet for data transfer efficiency. A/C (2021): Established scalable production with fabs. A/T (2021): Enhanced TeraPHY™ and introduced software tools.	G (2022): Targeted leadership in optical I/O for computing. A (2023): Invested in packaging technologies for optical I/O integration. T (2023): Supported emerging computing architectures. C (2024): Broadened applications to automotive and 5G. G (2024): Focused on sustainability in data centers and HPC.

[hyunjimoon]

decided to first focus on gene therapeutics and Intelligence Processing Unit (IPU) industry which has following features. I chose CRISPR Therapeutics, Editas Medicine, Bluebird Bio and Ayar Labs, SambaNova Systems, Mythic AI for action sequencing.

Gene Therapeutics:

• Targeted Treatment: Directly targets and modifies genes responsible for diseases, offering precise and potentially curative interventions.
• Diverse Delivery Systems: Utilizes viral vectors and non-viral methods to efficiently deliver therapeutic genes into patient cells.
• Permanent and Temporary Modifications: Offers both permanent corrections (e.g., CRISPR-Cas9) and temporary gene expression modulation (e.g., mRNA therapies).
• Broad Application Spectrum: Applicable to a wide range of diseases, from genetic disorders to certain cancers and viral infections.
• Regulatory and Ethical Considerations: Subject to strict regulatory and ethical scrutiny, especially concerning safety, efficacy, and germline modifications.

IPUs:

• Optimized Data Handling: IPUs are designed to handle the massive and complex data structures typical in machine learning, such as tensors, with high efficiency.
• Parallel Processing Capabilities: They often feature a large number of cores optimized for parallel processing, enabling them to perform many operations simultaneously, which is crucial for accelerating AI computations.
• Memory Architecture: IPUs typically have a unique memory architecture that allows for faster data access and processing, reducing bottlenecks that can occur in traditional computing systems.
• Software Ecosystem: The effectiveness of an IPU also depends on the software stack developed to leverage its hardware capabilities, including compilers, libraries, and tools that facilitate the development and deployment of AI models.

IPUs have challenges on:

1. Scalability Challenges: Both startups face unique scalability challenges due to their innovative approaches. SambaNova Systems must manage the integration of complex software and hardware, requiring advanced semiconductor manufacturing capabilities and a robust supply chain. Mythic AI, on the other hand, deals with the production of analog AI chips, a niche area that presents its own set of manufacturing and scalability challenges. Analyzing these companies can reveal how startups navigate the complexities of scaling up production for cutting-edge technologies.

2. Market Demand and Supply Chain Management: Both companies operate in the highly competitive AI market, where demand for efficient AI processing solutions is growing rapidly. Their capacity planning strategies likely include considerations for not just manufacturing scalability but also supply chain resilience, quality control, and the ability to adapt to changing market demands.

3. Partnerships and Infrastructure: Exploring how SambaNova and Mythic AI have developed partnerships with semiconductor fabs and managed their supply chains can provide insights into the strategic decisions startups must make to ensure their innovative products can be manufactured at scale and meet the market demand.

[hyunjimoon]

example of three startups from Gene Therapeutics industry

Company	Nail Phase	Scale Phase	Sail Phase
CRISPR Therapeutics	C (2013): Identified genetic disorders as targets. T (2013): Invested in CRISPR-Cas9 optimization. A (2014): Established partnerships. C (2015): Engaged with patient groups. T (2016): Demonstrated proof of concept.	A (2017): Partnered for gene therapy production. C (2018): Launched trials. T (2019): Expanded platform. A/C (2020): Strategic partnerships. A/T (2020): Ramped up facilities.	T/A (2021): Advanced to Phase III trials. C (2021): Entered CAR-T market. A (2022): Broadened collaborations. G (2023): Focused on global accessibility. T (2024): Continued CRISPR innovation.
Editas Medicine	A (2015): Collaboration with Juno Therapeutics. T (2015): Planned LCA10 clinical trial. C (2018): FDA approval for EDIT-101 IND.	A/T (2020): First in vivo CRISPR editing attempt. T (2020): Development of EDIT-301.	T (2022): Successful sickle cell therapy engraftment. A (2022): Created potency assay lifting clinical hold.
Bluebird Bio	A (2013): Collaboration with Celgene. T (2014): First patient treated in sickle cell and β-thalassemia studies. C (2016): PRIME designation by EMA.	A (2016): Strategic agreement with Lonza. T (2013): Acquisition of Pregenen. A/C (2018-2019): Strategic collaboration with Regeneron and EU approval for ZYNTEGLO™.	A (2018): Partnership with Gritstone Oncology, Inc. C (2018): Breakthrough Therapy Designation in ALD. T (2023): FDA approval for Lyfgenia.

example of three startups from IPU industry

Company	Nail Phase	Scale Phase	Sail Phase
Ayar Labs	C (2015): Identified bandwidth and power efficiency challenges. T (2015): Developed silicon photonics for optical I/O solutions. A (2016): Co-developed optical I/O with industry partners. C (2017): Aligned product development with market needs. T (2018): Integrated optical I/O with semiconductor platforms.	A (2019): Expanded R&D for optical I/O technology. C (2020): Piloted optical I/O with cloud providers. T (2020): Launched TeraPHY™ chiplet for data transfer efficiency. A/C (2021): Established scalable production with fabs. A/T (2021): Enhanced TeraPHY™ and introduced software tools.	G (2022): Targeted leadership in optical I/O for computing. A (2023): Invested in packaging technologies for optical I/O integration. T (2023): Supported emerging computing architectures. C (2024): Broadened applications to automotive and 5G. G (2024): Focused on sustainability in data centers and HPC.
SambaNova Systems	T (2017): Technology development and patent filings. A (2017): Establishing the company and initial R&D efforts.	A (2019-2022): Expansion and partnerships, like the development of DataScale system and international expansion, notably into the Japanese market. T (2020-2022): Product and technology development, including Dataflow-as-a-Service GPT Banking. C (2020-2022): Customer adoption and market validation, with notable partnerships for AI applications.	T (2022): Innovative product launches, like the introduction of the SN40L chip. G (2022-2024): Strategic focus on generative AI and resource reallocation towards high-demand AI applications. A (2022-2024): Continued expansion and adaptation to changing market demands.
Mythic AI	T (2012): Technology development and initial product launches, focusing on analog technology for chips. A (2012): Securing government contracts and initial R&D focus, including projects with the U.S. Air Force.	A (2012-2021): Funding rounds for scaling production, including over $170 million raised from investors for mass production and development of next-generation processors. T (2012-2021): Development of next-generation processors, such as the launch of the M1076 chip for broader edge AI applications. C (2012-2021): Broadening customer base to industries beyond defense, like the partnership with Lockheed Martin.	A (2023): Strategic funding for innovation and market expansion, with a $13 million round to bring its next-gen analog computing solution to market. T (2023): Continued technology advancement with a focus on analog computing, preparing to release its next-gen chip, the M2000. C (2023): Renewed focus on defense and expansion into public safety, industry, and consumer verticals.

[hyunjimoon]

Industry/Phase	Nail Phase	Scale Phase	Sail Phase
Gene Therapeutics	Technology Development (T): Strong positive correlation with Atom supply/architecture (A), focusing on establishing technical capabilities and foundational operational structures. Customer Development (C): Moderate positive correlations with A and T, highlighting initial market engagement and alignment.	Customer Development (C) & Atom Supply/Architecture (A): Strong positive correlation indicates an increased focus on aligning operational capabilities with market expansion. Technology/Organization (T): Critically linked with all actions, reflecting continued investment in product development and technological innovation.	Technology/Organization (T) & Atom Supply/Architecture (A): Strong positive correlation emphasizes the importance of technological innovation and advanced operational capabilities. Goal (G): Heightened correlations with T and A suggest a strategic emphasis on leveraging technological advancements and operational excellence.
IPU Industry	Foundation Establishment: Moderate positive correlations across the board, indicating interconnected initial efforts in technology development, market understanding, operational setup, and strategic goal alignment at a foundational level.	Operations & Market Engagement: Strengthened correlations between Atom supply/architecture (A) and Customer development (C), and across all actions with Goal (G), highlight the focus on expanding operations and deepening market engagement aligned with strategic goals for growth.	Technological Innovation: Strong correlations between Technology/organization (T) and Atom supply/architecture (A), and with Goal (G), suggest a focus on advancing product capabilities, optimizing operations, and aligning with strategic goals to explore new opportunities and sustain growth.

This table encapsulates the strategic and operational focus areas for both industries across different growth phases, emphasizing the evolving priorities and how various actions (A, G, C, T) interrelate to drive progress from establishment through scaling to adaptation and innovation.

[hyunjimoon]

이성환 대표님, Angie
started chatting by
SV 조인 (4년, sk 8년); vc 70% (펀드, 1억불, 데브라 박사)
company building, 좋은 기술 (스타트업 쿠킹,바이오텍)

전략 (미국, 한국 차이 - exit (60 MA, 30IPO) VS posdaq (vs nasdaq; 본인들이 judge심사, 실라젠; 안정적인 회사 상장; best in class (미국에 2억불), 파이프라인 실패 - 파이자 (auto immune, oncology, gene therapy; need planB, 먹튀방지 (창업자가 inventor))))

1. first in class 상장 어렵
2. 다양한 pipeline
3. inventor

모더나는 상장불가 (rna 신약 (비교불가), 하나만 팜, steven bangel mba (신약승인 15년 2b이상 자금 - 스토리))
신발 전문점, (best in class보단 first in class)

1. 스타트업 인수 어렵 (셀트리온, 종근장, 유한양행; 30조> 다케다는 40조 (10조를 R&D-capX))
2. moderna (nobar afeyan - seriesC까지 혼자 투자함 flagship (혼자함), 데릭로지 rna stem; 밥랭어-biologics (flagship lab), 더 메신저 (코비드 - 스티브 방셀, 백신만들까 말까; pivot (지카바이러스-백신을 가장 빨리 만듦)))

semiconductor (CMO) - fabless (nvidia)
moderna (정부로부터 1.5billion; manufacturing scaleup)
biotech 13, 15년 (2b,3b) - blockbuster약은 불확실

(경쟁자, price로 reimbersement안 해줌, )
paper를 보고 얼마나 성공 (전임상단계 - 타겟물질을 잘 만들어서 fda approval, delivery vehicle)

Alnylam 낸 약이 lnp (https://www.amvuttra.com/within-reach?gad_source=1&gclid=Cj0KCQjwir2xBhC_ARIsAMTXk86gwlSX3_bqzlxqggq_2PwwG6ScTrXVgTzQbrd7oM4W-VIP_LTSiCAaAoABEALw_wcB&gclsrc=aw.ds신약승인 - 검증된 )

TAM 완치약; rare disease patent cliff (roi)

1. funding - 약이 되면 michael j fox (파킨슨 - 부자 스폰서)
2. big pharma (파이프라인 - 나와 연결가능 clue; e.g. pfizer 전립선암); 2-3b; license in, out 연구같이 하다가 (blockbuster 1billion, 1년에 줘야할 돈이 1억불 (royalty 10% -); 1억불) - upfront

stem sell reprogramming (cancer vaccine)

• cancer market이 크다고 봤는데, (표적항암제, 화학항암제, 면역항암제 (몸이 암세포 죽이게 함 - 피 뽑아서)); 바이오텍은

보험사(reimbursement), fda, bigpharama) (semiconductor (customer가 big pharma (b2b); 의사,환자가 고객이 아님; )

genetech, 1937 (인슐린 대량생산); 캠브리지가 반도체로 리딩 (pc; fairchild, intel),휴스턴 (spacetech); 필링공장 (폴라로이드) 화학필림,

바이오텍 주가흐름 (30년 backtracking; 튀는해 sec quarterly ; 10k 사업보고서 다운, 키워드 뽑아서; 기업가치변동; market, product change; 암에서 rare event ())

new business plan (new market, product) -> reasoning (10k 상세히 주주들); public회사 (바이오텍)

[hyunjimoon]

Unifying statistical and pragmatic Linguistic Meaning in Term Sheets cld compares statistical and pragmatic meanings in the context of term sheets:

"In the context of term sheets, statistical meaning refers to the typical ranges, frequencies, and patterns of specific terms across a large sample of deals. It represents the quantitative norms and trends in how these terms are structured. Pragmatic meaning, on the other hand, reflects the functional implications, strategic intentions, and contextual interpretations of these terms. It encompasses how the terms are actually used and understood by the parties involved, often going beyond their literal definitions. For instance, the statistical meaning of an option pool might be its average size across deals, while its pragmatic meaning could include its role in talent attraction and retention strategies."

Option Pool:
Semiconductor: Typically larger (15-20%) due to need for highly specialized talent and faster product cycles. May include more frequent refreshes.
Life Science/Pharma: Often smaller initially (10-15%), but may grow over time. Longer development cycles allow for more gradual expansion.

Board of Directors:
Semiconductor: Often includes more technical experts. May have a balanced structure early on, shifting to investor-heavy later for faster decision-making.
Life Science/Pharma: Usually larger boards (5-7 members) with scientific and regulatory experts. Often VC-heavy due to capital intensity and long development timelines.

Vesting:
Semiconductor: Standard 4-year vesting with 1-year cliff. May include accelerated vesting tied to product launches or acquisitions.
Life Science/Pharma: Longer vesting periods (5-6 years) common. Often includes milestone-based vesting tied to clinical trial phases or regulatory approvals.

-wish to further understand https://techventures.columbia.edu/term-sheet-recommendations-for-launching-university-startups 's meaningful involvement

[hyunjimoon]

scoping both argument and target example using Analyzing Mokyr's Theory of Breakthrough Innovations cld

1. argument: Physical visibility → shared intuition → easier institutional alignment
2. target example: within mobility ventures (ev charging vs av driving software)

Mobility Ventures Industry Comparison

Aspect	EV Infrastructure 🔴💜	Autonomous Driving Software 🟩💜	Smart Parking Solutions 🟩🔴💜
Development	Physical hardware development for charging/swapping stations, power management systems, and mechanical components	Software development for perception, decision-making, and control systems; AI/ML model training	Mix of hardware (sensors, cameras, gates) and software (payment systems, space management)
Production	High initial costs for manufacturing facilities, ongoing material costs, supply chain management	Primary costs in computing infrastructure, data centers, and testing facilities	Medium hardware production costs, software deployment costs
Cost Structure/Scalability	High unit costs, requires significant capital for each new location; Linear scaling with physical infrastructure	Near-zero marginal cost for software deployment; Exponential scaling possible through OTA updates	Mixed scaling: Hardware components scale linearly, software scales exponentially
Distribution	Physical installation required at each location; Heavy logistics and maintenance needs	Digital distribution through OTA updates; Remote deployment possible	Hardware installation at parking locations plus cloud-based software deployment
Updates and Improvements	Physical components require on-site maintenance; Hardware upgrades need replacement	Continuous software updates possible; Rapid iteration through data feedback	Software updates deployed remotely; Hardware updates require physical intervention
Data Collection and Analytics	Limited to operational metrics unless IoT-enabled; Physical sensor data	Rich data collection from vehicles' sensors; Real-time performance monitoring	Comprehensive data from parking sensors, payment systems, and usage patterns
End of Life	Physical decommissioning required; Recycling/disposal considerations; Environmental impact	Software phase-out through updates; Server shutdown; Data archival	Mixed approach: Hardware removal/recycling plus software deprecation
Entry Barriers	High - Requires significant capital, real estate, and regulatory approval	Medium - Requires technical expertise but lower capital needs	Medium - Needs both hardware capability and software expertise
Cell Theory	Focus on specific charging/swapping solutions for different vehicle types	Specialized autonomous solutions for specific use cases (trucks, last-mile)	Niche solutions for different parking environments (street, garage, campus)
Ecosystem	Integration with power grid, vehicle manufacturers, property owners	Partnerships with vehicle OEMs, regulators, mapping providers	Connections with property managers, payment processors, cities
Evolution	Gradual improvements in charging speed, efficiency, and compatibility	Rapid advancement in AI capabilities and safety features	Iterative improvements in both hardware reliability and software features
Phylogeny	Evolved from traditional gas infrastructure model	Derived from robotics and computer vision advances	Emerged from traditional parking management systems
Properties of Life	Growth through physical expansion of network; Adaptation to new charging standards	Growth through software capability expansion; Adaptation to driving conditions	Growth through combined physical/digital expansion; Adaptation to urban needs

[hyunjimoon]

Here's a table illustrating the "Physical visibility → shared intuition → easier institutional alignment" pattern:

Analysis Steps	Physical/Visible Technology	Digital/Abstract Technology	Observable Institutional Impact
Historic Pattern	Steam Engine (1800s) • Visible safety risks (explosions) • Observable mechanical principles • Clear cause-effect in failures → Led to rapid boiler safety standards	Early Computer Networks (1970-90s) • Invisible data transmission • Abstract protocols • Complex failure modes → Took decades to develop internet standards	Physical: • Quick public consensus on safety needs • Clear standards within ~10 years • Consistent across regions Digital: • Fragmented understanding of risks • Standards took 20+ years • Regional variations persist
Modern Mobility Example	EV Charging Infrastructure • Visible charging connectors • Observable electrical safety issues • Clear physical compatibility needs → Rapid convergence on standards (CCS, CHAdeMO)	Autonomous Driving Software • Invisible decision algorithms • Abstract safety parameters • Complex edge cases → Still debating safety frameworks	Physical: • Clear charging standards by 2010s • Consistent safety protocols • High public trust Digital: • Ongoing safety framework debates • Regional regulatory differences • Lower public confidence

Key Pattern: When technology's workings and risks are physically visible:

1. Public intuitively understands the need for standards
2. Stakeholders more quickly align on safety requirements
3. Institutions can more rapidly develop clear frameworks
4. Regional harmonization happens faster

This supports your argument that physical visibility facilitates institutional alignment through shared intuitive understanding.