Peter Yang Launches Open-Source AI Skill to Fight Cancer Red Tape

Peter Yang

Artificial intelligence aids cancer patients far more by untangling fragmented administrative bureaucracy and tracking records than by analyzing dense clinical research.

AI-assisted summary of My Mom Has Cancer Again. Here’s How I’m Using AI to Fight It — original source. Check the original for full context.

Source published: 2026-11-04

Key ideas

Peter Yang developed an open-source ChatGPT skill named 'F-Cancer' to assist caregivers in navigating medical bureaucracy, generating one-page briefs, and tracking treatment options. (0:00)
Critical medical assets such as digital DICOM scans and physical biopsy pathology blocks often take several weeks to retrieve and should be requested immediately from hospital portals. (5:13)
Top-tier second opinions from Comprehensive Cancer Centers and clinical trial research can take up to a month to arrange, requiring parallel pursuit alongside standard-of-care treatments. (10:34)
Sharing a cancer journey publicly or across networks can unlock vital expert advice, which can be managed efficiently by distributing AI-generated email updates to an informal board of advisors. (14:19)

Detailed notes

## Background Published on YouTube by Peter Yang on November 4, 2026, this video addresses how caregivers can use AI to navigate a loved one's cancer diagnosis following his mother's fourth recurrence. According to the video description, Yang introduces an open-source ChatGPT skill he built called 'F-Cancer' to help families organize medical records, research therapies, consult experts, and manage the overwhelming logistics of care.

## TL;DR Drawing on his family's personal experience, Peter Yang details a five-step playbook for caregivers confronting cancer: building a centralized one-page patient brief, hunting down fragmented records and physical pathology specimens, and benchmarking standard care against clinical trials and tumor sequencing options. He explains how to leverage AI to draft concise communications for second opinions and assemble an informal expert advisory board by sharing the diagnosis openly. Throughout the breakdown, Yang emphasizes that automating administrative friction is essential not just for medical advocacy, but to protect precious, finite time with loved ones.

## Atmosphere & style The presentation carries a quiet, vulnerable, and deeply earnest tone that blends the emotional gravity of a child caring for an aging parent with the structured problem-solving of a systems engineer. Yang speaks candidly about the exhausting gap between hyped medical breakthroughs and the slow, fax-dependent bureaucracy of real-world hospitals. The delivery shifts effortlessly from pragmatic screen demonstrations to tender personal reflections on mortality and family, leaving the audience feeling both grounded and empowered to take action.

## Bulletin - [00:00](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=0s) Peter Yang developed an open-source ChatGPT skill named 'F-Cancer' to assist caregivers in navigating medical bureaucracy, generating one-page briefs, and tracking treatment options. - [05:13](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=313s) Critical medical assets such as digital DICOM scans and physical biopsy pathology blocks often take several weeks to retrieve and should be requested immediately from hospital portals. - [10:34](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=634s) Top-tier second opinions from Comprehensive Cancer Centers and clinical trial research can take up to a month to arrange, requiring parallel pursuit alongside standard-of-care treatments. - [14:19](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=859s) Sharing a cancer journey publicly or across networks can unlock vital expert advice, which can be managed efficiently by distributing AI-generated email updates to an informal board of advisors.

## Key takeaways - [00:52](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=52s) Facing cancer impacts nearly every family, with approximately two in five Americans receiving a diagnosis in their lifetime and 81% of diagnoses occurring at age 55 or older. - [02:04](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=124s) AI is often most valuable to caregivers not for reading theoretical research papers, but for managing the logistical chaos of doctor appointments, follow-ups, and fragmented records. - [02:04](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=124s) The custom F-Cancer skill consolidates patient information into a structured one-page Google Doc covering contacts, next actions, clinical facts, a plain-English glossary, and a care log. - [02:04](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=124s) Generating the top three priority questions via AI prior to tight 15-to-20-minute clinical appointments maximizes face-to-face time with oncologists. - [05:13](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=313s) AI recommendations are only as reliable as the underlying records, requiring immediate collection of written summaries, raw DICOM imaging, and physical pathology blocks. - [05:13](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=313s) Physical pathology tissue blocks are the single most critical asset for genomic sequencing and accurate second opinions, but often take weeks to retrieve from off-site storage. - [07:41](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=461s) A massive operational disconnect exists between cutting-edge cancer research and the archaic reality of clinical workflows reliant on faxes and voicemails. - [07:41](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=461s) Caregivers can understand first-line therapy expectations by referencing publicly available National Comprehensive Cancer Network (NCCN) guidelines with AI assistance. - [10:34](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=634s) Securing second opinions from premier Comprehensive Cancer Centers like MD Anderson, MSK, or UCSF often takes up to a month and requires substantial preparation. - [10:34](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=634s) Comprehensive tumor sequencing through services like BostonGene identifies actionable DNA and RNA biomarkers that traditional pathology may overlook. - [12:47](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=767s) Caregivers should evaluate clinical trials concurrently with initial care via ClinicalTrials.gov, while exercising caution regarding early unproven phase 1 and 2 trials. - [14:19](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=859s) Openly sharing a cancer diagnosis rather than keeping it secret can mobilize valuable advocates, clinicians, and specialized organizational support. - [14:19](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=859s) ChatGPT can translate dense clinical reports into concise bulleted updates, allowing caregivers to manage an external advisory board efficiently via BCC emails. - [15:41](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=941s) Beyond medical navigation, caregivers must deliberately protect quality time with loved family members before career ambitions and busywork displace irreplaceable moments.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=0s) My Mom's Story and How AI Can Help

Peter Yang begins by sharing that his mother has survived cancer three times, but she has recently been diagnosed yet again. During the prior three battles, his parents carried most of the burden alone, but he resolves that he will not simply watch from the sidelines this time: “The first three times this happened, my parents carried most of the burden. But this time, I'm not going to be a bystander.” Yang views supporting an ill family member as “probably the most important AI use case of them all.” To help patients and their families handle the immense logistical and emotional load, he developed an open-source, free ChatGPT skill named **F-Cancer** (hosted at `github.com/petergyang/fuck-cancer`), designed to guide caregivers step by step through the healthcare maze.

[00:52](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=52s) Cancer Statistics and the Caregiver's Burden

Cancer remains an inescapable reality across modern society. In the United States, roughly two in five people will receive a cancer diagnosis at some point in their lifetime, meaning that once caregivers and immediate family members are accounted for, virtually every household is impacted. On a positive note, cancer mortality rates have declined by 35% since 1991 thanks to ongoing advances in medical technology and therapies. Despite these improvements, age remains a major risk factor; approximately 81% of all cancer diagnoses occur in individuals aged 55 and older.

Facing this diagnosis requires a structured plan, which Yang divides into five distinct steps for caregivers and patients: 1. Create a one-page brief summarizing up-to-date facts and logistical contacts. 2. Track down and organize all medical records, lab reports, and imaging scans. 3. Thoroughly research standard treatment paths and clinical options. 4. Seek second opinions and bring experienced specialists into your corner. 5. Fiercely protect quality time with family.

[02:04](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=124s) Step 1: Create a One-Page Brief with AI

When his mother’s cancer recurred, Yang initially assumed he would leverage AI primarily to read dense oncology studies. In reality, he discovered that the tool's greatest utility lies in administrative coordination: “Instead, it's been far more useful for staying on top of the mountain of paperwork and follow-ups across the different doctors and hospital departments.” To solve this administrative chaos, the F-Cancer skill builds and maintains a live, concise Google Docs brief organized into five key components:

- **Patient Information:** Baseline demographics, emergency contacts, primary care team members, second-opinion contacts, phone numbers, and email addresses. - **What to Do Next:** The top three near-term action items or the top three priority questions to raise with doctors. - **What We Know:** Concrete clinical facts—biomarkers (such as ER/PR status), cancer stage, current treatment status, and diagnosis specifics. - **Plain-English Glossary:** Clarifications and translations of complex medical jargon into easy-to-understand language. - **Care Log:** A dated, chronological journal tracking appointments, test results, and newly discovered details.

```mermaid flowchart TD accTitle: Five Sections of the F-Cancer One-Page Brief accDescr: Diagram showing the five core sections built into the Google Docs family brief via the F-Cancer AI skill.

subgraph BriefStructure["One-Page Cancer Brief"] A["1. Patient Information & Care Team"] B["2. What to Do Next (Top 3 Questions)"] C["3. What We Know (Stage, Biomarkers)"] D["4. Plain-English Medical Glossary"] E["5. Chronological Care Log"] end

A --> B B --> C C --> D D --> E ```

To demonstrate the setup, Yang walks through a ChatGPT prompt: installing the skill directly from GitHub, then submitting a prompt such as: `F Cancer, my mom was just diagnosed with stage three breast cancer. Here are her markers, ER and PR positive... and here's her oncologist.` The tool outputs the structured brief directly. Having this single document ready to share with family members has proved invaluable when walking into tight 15-to-20-minute clinical visits. Caregivers can prompt the model just before an appointment—asking it to pull the top three most urgent questions based on the brief—ensuring precious consultation time is not wasted.

[05:13](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=313s) Step 2: Organize Medical Records and Scans

An AI prompt is only as reliable as the source records behind it: “Your brief is really only as good as the data behind it. So, the more records you have on the patient, the more personalized and up-to-date AI's responses can be.” Because his mother was treated at several facilities over the years, her medical history was fragmented across disparate hospital patient portals. Looking back, Yang strongly advises families to request three distinct categories of medical assets immediately rather than waiting:

- **Written Records:** Prior pathology summaries, laboratory panels, surgery summaries, scan impressions, and full treatment histories. - **Digital Imaging Files:** Raw CT, PET, and MRI scans. These are typically stored in the specialized **DICOM** imaging format, take two to three weeks to process, and are often delivered physically on a USB thumb drive or CD-ROM. - **Pathology Blocks and Slides:** Tissue specimens preserved from the tumor biopsy. This physical tissue is the single most critical asset for genomic sequencing and accurate second opinions. Because facilities often store old specimens in off-site warehouses, tracking them down takes the longest amount of time; his mother is still waiting on blocks from a previous cancer.

Yang stores all gathered records inside a shared Google Drive. Connecting ChatGPT directly to this archive alongside the one-page brief grounds the AI's responses in the patient’s personal medical reality rather than general web advice. Users can also ask ChatGPT to streamline the record retrieval process itself; for example, prompting it with hospital names like Stanford or UCSF will generate direct request portal links, estimate fulfillment wait times, and even draft initial record request emails.

[07:41](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=461s) Step 3: Research Treatment Options and Navigate Bureaucracy

A patient’s diagnosis typically begins when an abnormality is detected on an initial scan (such as a CT scan), followed by a confirmatory biopsy. Afterward, a multidisciplinary **tumor board**—consisting of oncologists, radiologists, surgical oncologists, and other specialists—meets to deliberate over the case and propose standard-of-care treatments based on established national protocols (like the US NCCN guidelines), which generally involve surgery, chemotherapy, radiation, or immunotherapy.

However, this formal clinical pipeline moves agonizingly slowly. Two full months after his mother's initial suspicious CT scan, her family was still waiting to confirm whether the mass was a recurrence of her previous breast cancer or a brand-new primary cancer. Yang highlights the deep emotional disconnect between online hype and the archaic reality of day-to-day healthcare: “Working with my mom's medical team has sometimes felt like working with a co-worker that doesn't respond to emails or calls and only communicates through voicemails and faxes.” He expresses deep frustration with tech commentary hyping miraculous AI cures while families are stuck waiting on physical mail and fax machines:

> “So, there's a massive disconnect between what state of the art treatment looks like and what standard care and procedure looks like. And as a caregiver, I think it's my job to really fight for my mom and try to get her the best care possible and navigate a lot of the bureaucracy.”

Rather than waiting passively while fearing the tumor is growing, caregivers can take concrete steps. The first move is understanding the baseline standard of care by consulting the National Comprehensive Cancer Network (NCCN) guidelines, which are publicly available with a free account. Caregivers can feed their patient brief and pathology details to AI models or specialized tools to surface the expected first-line therapies and prepare informed questions.

[10:34](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=634s) Researching Standard Care, Second Opinions, and Tumor Sequencing

Navigating cancer treatment begins with establishing an evidence-based benchmark for standard care and seeking out expert second opinions. Using his customized ChatGPT skill, Peter Yang demonstrates how a caregiver can prompt the model with specific questions: "What does standard treatment look like for a case like the patient's? And where do we get a second opinion?" In the hypothetical case of hormone receptor-positive breast cancer (ER and PR positive), the AI lays out typical surgical approaches and hormone-blocking therapies matched to the patient’s exact biomarkers. Yang emphasizes that understanding standard care equips families to have informed, reasoned discussions with their oncology team. For second opinions, he strongly recommends seeking consultations from top-tier Comprehensive Cancer Centers, citing institutions such as MD Anderson in Texas, Memorial Sloan Kettering (MSK) in New York, Dana-Farber Cancer Institute in Boston, and Stanford or UCSF in the Bay Area.

Securing a second opinion, however, is a grueling administrative undertaking that can easily take up to a month. Yang shares that his family is currently in the middle of securing an opinion from UCSF, and the legwork done in Step 2—organizing scans, pathology reports, and clinical records—has proven vital. He actively uses AI to parse medical files, draft formal cover letters explaining his mother's clinical history, and compose outbound emails to intake coordinators. Alongside second opinions, Yang underscores the necessity of comprehensive tumor sequencing. Because cancer is biologically heterogeneous, molecular profiling offers insights that standard pathology might miss. He outlines his plan to use BostonGene to run comprehensive DNA, RNA, and biomarker sequencing at an out-of-pocket cost of roughly $4,000, which yields a detailed molecular report capable of pinpointing targeted therapies.

As Yang explains regarding the individuality of the illness: “Cancer is actually more than 100 different types of diseases, and each person's tumor can behave differently.”

- **Recommended Cancer Centers**: MD Anderson (Texas), Memorial Sloan Kettering (New York), Dana-Farber (Boston), Stanford Medicine (California), and UCSF Health (California). - **Turnaround Time for Second Opinions**: Up to one full month from initial intake to clinical review. - **Tumor Profiling Service**: BostonGene, costing approximately $4,000 for in-depth DNA/RNA sequencing to identify actionable biomarkers.

```mermaid flowchart TD accTitle: Cancer Research and Expert Workflow accDescr: Visual flowchart outlining the parallel tracks for researching treatment, obtaining sequencing, and leveraging an expert network. subgraph S1["Step 3: Clinical Research"] A["Patient Records & AI Brief"] --> B["Identify Standard Care & Biomarkers"] A --> C["Submit for Second Opinion<br/>(MD Anderson, MSK, UCSF, Stanford)"] A --> D["Comprehensive Tumor Sequencing<br/>(BostonGene DNA/RNA profiling)"] A --> E["Search ClinicalTrials.gov<br/>(Targeted Phase 2/3 trials)"] end

subgraph S2["Step 4: Expert Advisory"] B & C & D & E --> F["Synthesize Key Medical Updates"] F --> G["Draft Oncologist Email Updates via ChatGPT"] G --> H["BCC Expert Board<br/>(Advocates, Foundation staff, Clinicians)"] H --> I["Actionable Feedback & Treatment Strategy"] end ```

---

[12:47](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=767s) Evaluating Clinical Trials with AI and ClinicalTrials.gov

When a patient receives an initial diagnosis, oncologists reflexively start with standard of care because those regimens are rigorously tested and clinically proven. However, cutting-edge therapies are often locked within clinical trials. Yang points out that investigating trials is a critical parallel path, though it requires significant caution. Early-stage research presents real hazards, meaning patients cannot jump into experimental protocols blindly without seasoned clinical guidance.

Because the AI skill retains the complete context of the patient's one-page medical brief and underlying records, it can query ClinicalTrials.gov—the central registry of US clinical studies—to extract trials specifically matching the patient's diagnosis, stage, biomarker profile, and geographical location. When prompted to identify the top three clinical trial candidates across the United States or Canada, the model returns tailored options within driving distance of the patient. Yang stresses that researching trials, requesting second opinions, and awaiting sequencing results should never be deferred until standard therapies fail; they take weeks or months to arrange and should be pursued early in the disease course.

Yang issues a direct warning about jumping too hastily into experimental regimens: “And by the way, you should be very careful about clinical trials because some stage one and two trials are actually unproven and could actually make things worse.”

- **Database Utilized**: ClinicalTrials.gov, integrated directly into the custom AI skill to match inclusion/exclusion criteria. - **Trial Screening Parameters**: Phase, geographic proximity to the patient, prior therapy requirements, and specific tumor histology. - **Strategic Advice**: Initiate clinical trial matching concurrently with initial treatment planning rather than waiting for disease recurrence.

---

[14:19](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=859s) Step 4: Assembling an Expert "Board" and Sharing the Journey

A cancer diagnosis often drives families into secrecy, but Yang found that making his mother's situation public unlocked an outpouring of specialized assistance. Over a dozen individuals reached out with firsthand advice, resources, and connections. In particular, he highlights three invaluable allies who stepped into his corner: Ari from OpenCancer, who connected the family with specialists for pathology reviews and advanced diagnostic testing; Bailey from the Sijbrandij Foundation, who helped structure precise questions regarding tissue sequencing and therapy pathways; and practicing clinicians such as Dr. Odabashian, who reached out via direct messages on X (formerly Twitter) to offer independent clinical perspectives.

To maintain this support network without drowning in communication logistics, Yang treats these advisors as an informal board of directors. Whenever fresh pathology arrives or a treatment crossroad emerges, he prompts ChatGPT to transform dense clinical notes into succinct, clear email updates: "Can you draft a short email update?" He then sends these summaries via BCC to his advisory panel, soliciting their feedback on proposed regimens, surgical margins, or sequencing targets. This crowdsourced clinical brain trust gives the family confidence that no stone is left unturned.

Yang reflects on why breaking the silence around illness is so transformative: “You know, most families who get cancer like to keep it to themselves. But, what I found is that sharing the story and sharing the journey online or even just with friends really empowers other people to come and help.”

- **Key Contributors**: - Ari (OpenCancer): Provided direct introductions for second-opinion pathology reviews and advanced lab testing. - Bailey (Sijbrandij Foundation): Coached the family on drafting critical questions for treating oncologists regarding tissue assays. - Dr. Odabashian and social media clinicians: Provided rapid independent perspectives on diagnostic findings. - **Communication Workflow**: Drafting concise clinical bullet-point memos with ChatGPT, then emailing the distributed expert board via BCC.

---

[15:41](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=941s) Step 5: Protecting Precious Time with Family

The final step Yang presents departs from medical administration to focus on the emotional reality of caregiving: preserving meaningful time with loved ones. Upon receiving news of his mother's recurrence, Yang stepped away from his routine and spent an entire uninterrupted week with her, walking together, sharing meals, and having long conversations. He also instituted a daily phone call so his parents can see and talk with his young daughters.

Yang observes that for professionals in their early forties, life becomes a complex squeeze between raising young children and supporting aging parents. At this career stage, ambition often pulls people toward status, promotions, and income. Yet, financial success ultimately proves that time is far more scarce than money. Because moments with family lack explicit deadlines and cannot be tracked on a balance sheet, they are the easiest commitments to postpone until it is too late. Referencing the *Beardo* comic strip by Dan Dougherty—which depicts a father holding his baby girl, watching her grow up frame by frame, and suddenly realizing she has moved out while he has grown old—Yang reflects that his own daughters are rapidly growing up, and life moves relentlessly fast.

He urges viewers to keep their core priorities straight: “I think if there's one thing that I made you realize in this video is to recognize that time with your spouse, your kids, and your parents is very precious and limited and can pass by before you know it, right?”

Yang concludes by reminding viewers that his open-source "F-Cancer" ChatGPT skill is freely accessible via GitHub in the video description. He encourages anyone confronting a cancer diagnosis—either personally or as a family caregiver—to reach out for support, share their stories openly, and lean on the community.

## Notable quotes > [14:19](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=859s) Most families who get cancer like to keep it to themselves. But, what I found is that sharing the story and sharing the journey online or even just with friends really empowers other people to come and help. — Peter Yang

> [12:47](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=767s) You should be very careful about clinical trials because some stage one and two trials are actually unproven and could actually make things worse. — Peter Yang

> [05:13](https://www.youtube.com/watch?v=Ddc5NFNnCfE&t=313s) Your brief is really only as good as the data behind it. So, the more records you have on the patient, the more personalized and up-to-date AI's responses can be. — Peter Yang

## Mentioned Peter Yang, ChatGPT, F-Cancer, Google Docs, Google Drive, NCCN, MD Anderson, Memorial Sloan Kettering, Dana-Farber Cancer Institute, Stanford Medicine, UCSF Health, BostonGene, ClinicalTrials.gov, Ari, OpenCancer, Bailey, Sijbrandij Foundation, Dr. Odabashian, X, Dan Dougherty, Beardo

## Video description Copy my best AI workflows to save time and automate busywork: https://www.behindthecraft.com/?utm_source=youtube&utm_medium=tutorial 🤝 Subscribe to my practical AI newsletter read by 150K+ professionals: https://creatoreconomy.so

My mom has survived cancer three times, and now she has it again. This time, I'm not going to be a bystander. In this video, I walk through the 5 steps I'm taking to help her navigate cancer with ChatGPT, using a free skill I built that any patient or caregiver can use. I cover how to create a one-page brief, organize medical records, research treatment options, find experts, and protect time with family. I'm not a doctor, and none of this replaces a care team.

What I cover: (00:00) My mom's story and how I'm helping (00:52) Why cancer touches almost every family (02:04) Step 1: Create a one-page brief with AI (05:13) Step 2: Organize medical records and scans (07:41) Step 3: Research treatment options, second opinions, and trials (14:19) Step 4: Find experts to have in your corner (15:41) Step 5: Protect quality time with family

Get my free F-Cancer skill: https://github.com/petergyang/fuck-cancer

Resources mentioned: OpenCancer: https://opencancer.ai Future of Cancer Care Today (Sijbrandij Foundation): https://sijbrandijfoundation.org/fcct ClinicalTrials.gov: https://clinicaltrials.gov Cancer stats (NCI SEER): https://seer.cancer.gov/statfacts/html/all.html Time with family data (Our World in Data): https://ourworldindata.org/grapher/time-spent-with-relationships-by-age-us Comic: Beardo by Dan Dougherty

Where to find me: X: https://x.com/petergyang LinkedIn: https://www.linkedin.com/in/petergyang/

Subscribe to this channel - more tutorials coming soon!

[Source](https://www.youtube.com/watch?v=Ddc5NFNnCfE)

中文简报

Peter Yang推出开源AI技能应对癌症繁琐行政流程人工智能在理顺分散的行政繁文缛节和追踪病历记录方面为癌症患者提供的帮助，远多于分析深奥的临床研究。结合其家庭切身体验，Peter Yang为面临癌症的照护者详述了一套五步操作指南：建立集中的单页患者摘要，追查零散的病历记录与实体病理标本，以及将标准治疗与临床试验及肿瘤测序方案进行对比基准分析。他说明了如何借助人工智能起草用于寻求第二意见的简要沟通文书，并通过公开分享确诊情况组建非正式专家顾问委员会。在整套方案解析中，Yang强调自动化处理行政障碍至关重要，这不仅关乎患者的医疗权益，更是为了守护与挚爱亲人共度的珍贵而有限的时光。 Peter Yang开发了一款名为“F-Cancer”的开源ChatGPT技能，以协助照护者应对繁复的医疗行政流程、生成单页摘要并追踪治疗方案。 (0:00) 诸如数字DICOM扫描影像和实体活检病理切片等关键医疗资产往往需要数周才能调取，应立即通过医院门户申请获取。 (5:13) 来自综合癌症中心的顶级第二诊疗意见以及临床试验研究所需安排时间可达一个月，因此需要在推进标准治疗方案的同时同步进行。 (10:34) 公开或在社交圈内分享抗癌经历能够获得至关重要的专家建议，而通过向非正式顾问委员会分发由AI生成的电子邮件动态，可以高效管理这些沟通。 (14:19)

Reading summary

Peter Yang outlines a practical framework using a customized, open-source AI skill to manage the administrative burden of cancer care, arguing that artificial intelligence is most effectively used to organize records, benchmark standard treatments, and coordinate specialist opinions. AI for Healthcare Coordination and Patient Briefs (2:04) Cancer touches roughly two in five Americans in their lifetime, with 81% of diagnoses occurring in people aged 55 and older. While cancer mortality has declined by 35% since 1991, families navigating care still face an overwhelming administrative burden. Yang notes that hospital communication often resembles an unresponsive coworker relying on faxes and voicemails rather than modern digital coordination. To alleviate this, he developed an open-source ChatGPT skill called 'F-Cancer' that compiles and maintains a live one-page brief. Rather than reading dense oncology literature, Yang found AI most useful for administrative organization. The brief structures five vital areas: basic patient and care team contacts, the top three immediate action items, confirmed clinical facts and biomarkers, a plain-English glossary of medical terminology, and a chronological care log. Caregivers can generate high-priority questions right before brief 15- to 20-minute oncology appointments to make the most of limited consultation time. Centralizing Records and Diagnostic Assets (5:13) AI outputs are only as accurate as the patient records supporting them. Because cancer treatment is often fragmented across multiple hospital systems, Yang advises caregivers to immediately gather three critical asset classes: written records (such as pathology reports, surgery summaries, and lab panels), digital imaging files (raw CT, PET, and MRI scans in DICOM format, often delivered on USB drives or CDs), and physical pathology blocks and slides. Tissue specimens from biopsies are the hardest to retrieve because hospitals frequently store them in off-site archives. However, physical blocks are necessary for accurate second-opinion reviews and comprehensive DNA and RNA sequencing. Yang notes that comprehensive tumor profiling—such as through BostonGene, which costs approximately $4,000 out-of-pocket—helps identify targeted therapies by analyzing the unique molecular behavior of individual tumors. Storing these files in a shared drive allows AI models to ground their analysis in the patient's specific clinical history. Parallel Research, Second Opinions, and Trials (10:34) Formal clinical workflows move slowly; Yang's mother waited two months after a suspicious CT scan just to establish whether a mass was a recurrence or a new primary cancer. To prevent passive waiting, Yang recommends that caregivers benchmark expected therapies against National Comprehensive Cancer Network (NCCN) guidelines and immediately seek second opinions from major Comprehensive Cancer Centers, such as MD Anderson, Memorial Sloan Kettering, Dana-Farber, Stanford, or UCSF, which can take up to a month to process. Caregivers can also prompt AI models integrated with ClinicalTrials.gov to screen for experimental protocols matching a patient's exact stage, histology, and location. However, Yang cautions that Phase 1 and Phase 2 trials carry genuine risks and may even worsen outcomes, meaning experimental regimens must be evaluated carefully with clinicians. He stresses that second opinions, tissue sequencing, and trial matching should be initiated in parallel with first-line treatment rather than deferred until standard care fails. Building an Expert Board and Preserving Time (14:19) Although illness often leads families to withdraw, Yang found that openly sharing his mother's diagnosis unlocked critical guidance from patient advocates, nonprofit foundations, and independent clinicians. Treating these contacts as an informal board of directors, he uses ChatGPT to turn dense clinical updates into clear summary memos sent via BCC to solicit feedback on treatment paths. Beyond administrative tactics, Yang underscores the necessity of fiercely protecting quality time with family. For professionals in their early forties who are balancing demanding careers and young children, non-urgent moments with aging parents are easily postponed until it is too late. He advocates stepping away from routine work to spend focused, everyday time with family while navigating the illness. While generative AI cannot replace clinical expertise or resolve institutional healthcare friction, using structured tools to organize medical records and pursue opinions in parallel helps caregivers advocate effectively while protecting irreplaceable time with loved ones.

精炼阅读摘要

Peter Yang 在其母亲第四次确诊癌症后指出，照料重病家人是人工智能最具价值的实际应用场景；他提出了一套结合开源工具、数据整理与专家协作的五步应对框架，帮助照料者突破医疗系统的行政延误并争取最佳治疗方案。利用AI精简行政沟通与问诊准备 (2:04) 面对复杂的医疗流程，Peter Yang 发现人工智能的核心价值并非直接研读高深医学文献，而是协助处理跨医院与科室产生的庞大文书及随访工作。为此，他在 GitHub 上发布了名为 F-Cancer 的免费开源 ChatGPT 技能，用于在 Google Docs 中生成并实时维护单页患者备忘录。该备忘录整合了五项关键内容：患者基本信息与多学科医疗团队联系方式、近期三大行动项或问诊问题清单、分期及生物标志物（如 ER/PR 状态）等确诊事实、复杂医学术语的通俗解释，以及按时间排序的就诊和检查日志。在日常门诊中，患者与肿瘤医生的交流时间通常仅有15至20分钟。照料者可以在候诊时基于该单页备忘录向 AI 提问，快速梳理出本次会诊最紧迫的三个问题，确保有限的交流时间聚焦于关键决策。备忘录还便于在家庭成员之间同步最新进展，降低因沟通脱节导致的决策损耗。系统化归集病历、影像与病理切片 (5:13) Yang 强调，AI 模型的回答质量完全取决于输入数据的完整度与准确性。由于其母亲多年来在多家医疗机构就医，病史高度碎片化，他建议家属在确诊之初立即调取三类医疗资产，而非被动等待：包括病理总结、化验单、手术记录及既往治疗史在内的书面报告；以 DICOM 格式存储的 CT、PET 和 MRI 等原始数字影像文件（通常需耗时两至三周刻录在光盘或 U 盘中递送）；以及肿瘤活检保留的病理组织块与切片。其中，物理病理切片通常存放在异地仓库中，调取周期最长，却是后续基因测序和精确二次诊疗不可或缺的基础。照料者可将收集到的所有资料归档至云端共享文件夹并关联 AI 工具，使模型输出能够基于患者真实的临床背景。用户还能利用 AI 自动查询 Stanford 或 UCSF 等机构的病历调取系统，获取预计处理周期，并由 AI 代写正式的档案申请邮件以加快进度。突破医疗官僚阻滞与多轨推进治疗研究 (7:41) 虽然美国癌症死亡率自1991年以来下降了35%，但约81%的癌症患者年龄在55岁以上，常规诊疗往往面临漫长行政审批的挑战。Yang 谈到，其母亲在发现 CT 异常后长达两个月仍在等待确认病灶属于旧疾复发还是原发新肿瘤；传统医疗团队在沟通中仍过度依赖语音留言与传真机，与公众对顶尖医疗技术的预期形成巨大落差。为了对抗这种被动等待，照料者需要自主通过美国国家综合癌症网络（NCCN）指南核实标准治疗路径，并借助 AI 依据生物标志物评估外科手术或激素阻滞疗法。此外，Yang 建议将第二诊疗意见、肿瘤分子测序与临床试验筛选等工作前置，与初始治疗规划同步推进，而非留待一线方案失败之后。获取 MD Anderson、MSK、Dana-Farber、Stanford 或 UCSF 等顶级综合癌症中心的专家意见往往耗时达一个月；自费约4000美元使用 BostonGene 等机构开展全面的 DNA/RNA 测序，可揭示常规病理遗漏的靶向治疗靶点。在利用 AI 对接 ClinicalTrials.gov 筛选相符的二期或三期临床试验时，Yang 特别提醒早期一期与二期试验存在未经验证的临床风险，照料者必须在专业医生指导下审慎评估，切忌盲目尝试。公开抗癌历程与组建分布式专家智囊 (14:19) 针对许多家庭在患病后倾向于隐瞒病情的心理，Yang 认为主动公开母亲的抗癌历程能引入关键的外部援助。通过社交媒体与公开渠道，他结识了协助对接病理复核的 OpenCancer 负责人 Ari、指导设计组织测序提问策略的 Sijbrandij 基金会成员 Bailey，以及在 X 平台提供独立临床视角的现任执业医生。这批跨领域的专业人士构成了一个非正式的“外部董事会”。为避免陷入繁复的沟通协调，Yang 将 ChatGPT 作为沟通秘书：每当获得新的病理数据或面临治疗路径选择时，利用 AI 将冗长的临床记录提炼成简明摘要，再通过密送（BCC）方式批量发给专家智囊团征求意见。该流程既维护了与外部支持者的联系，又为患者家庭在手术方案及药物靶点选择上提供了多维度的同行评审。回归照料核心：防范时间稀缺与守护家庭 (15:41) 在应对繁琐的医疗行政之外，Yang 提醒照料者必须全力守护与家人相处的质量。在得知母亲复发后，他主动暂停日常事务，全职陪伴母亲一周，共同散步与长谈，并建立了让年幼女儿与祖父母每日视频通话的习惯。他观察到，处于四十岁上下的人群往往面临抚养子女与照顾老人的双重夹击，在追求职业地位与收入的同时，极易将没有明确截止日期的家庭时光不断推迟。援引漫画《Beardo》对时光飞逝的刻画，Yang 警示读者金钱无法弥补时间的流逝，呼吁照料者在理性运用技术与繁琐体制周旋的同时，始终牢记生命的紧迫性，主动向外界社群寻求支持，将有限的时间留给最重要的家庭陪伴。 Yang 提供的整套方案展示了生成式 AI 作为个人医疗行政助手与研究工具的潜力，但也凸显出底层医疗体系对物理介质和漫长流程的依赖。尽管开源技能有助于平抑信息不对称，寻找前沿疗法与早期临床试验的真实风险依然需要高度依赖专业医生的临床研判与严谨求证。

Source quotes

[14:19] Peter Yang: Most families who get cancer like to keep it to themselves. But, what I found is that sharing the story and sharing the journey online or even just with friends really empowers other people to come and help. [12:47] Peter Yang: You should be very careful about clinical trials because some stage one and two trials are actually unproven and could actually make things worse. [5:13] Peter Yang: Your brief is really only as good as the data behind it. So, the more records you have on the patient, the more personalized and up-to-date AI's responses can be.

Original transcript

[0:00] My mom has survived cancer three times [0:03] and now unfortunately she has it again. [0:06] The first three times this happened, my [0:08] parents carried most of the burden. [0:10] But this time, I'm not going to be a [0:12] bystander. In this video, I want to walk [0:14] through five steps that I'm taking to [0:16] help my mom fight cancer with AI's help. [0:20] You know, caring for a loved one is [0:22] probably the most important AI use case [0:24] of them all. So, I want to show you [0:26] exactly how I do it using a free skill [0:28] that I built that any patient or a [0:31] caregiver can use. The skill is called F [0:34] Cancer and here's a quick clip of how it [0:36] works. [0:53] Okay, so before I show you how to use my [0:54] skill, here are some quick stats about [0:57] cancer. [0:58] Two in five people in the US will get it [1:02] at some point in their life. [1:04] When you include caregivers, that means [1:06] that almost no family will be left [1:09] untouched by this disease. [1:11] Now, the good news is that cancer death [1:12] rates have fallen by 35% since 1991 as [1:17] technology and treatment has progressed. [1:20] But many people will still get it when [1:21] they're older. About 81% of cases are in [1:25] people who are aged 55 or above. [1:29] So, how do you deal with this terrible [1:30] disease as a patient or a caregiver? And [1:34] where can AI help? [1:36] Well, I've spent a lot of time on this [1:39] and I think there's five steps that you [1:40] should follow. [1:42] Number one, create a one-page brief [1:45] with the latest facts and information. [1:47] Number two, organize all of the [1:49] patient's medical records and scans. [1:52] Number three and four, research [1:54] treatment options and find experts to [1:57] have in your corner. And perhaps most [1:59] important of all, protect quality time [2:03] with your family. [2:05] Okay, so the first step is to create a [2:07] one-page brief. [2:09] When my mom's cancer came back, I [2:10] thought I'd mostly use AI to research [2:13] her disease. [2:14] Instead, it's been far more useful for [2:16] staying on top of the mountain of [2:18] paperwork and follow-ups across the [2:21] different doctors and hospital [2:22] departments. [2:24] That's why I built my F Cancer Skill to [2:27] maintain a concise fact-based brief in [2:30] Google Docs. [2:32] And the first time you install the [2:33] skill, it's going to walk you through a [2:35] filling out this brief in five key [2:37] sections. So, number one, the patient [2:39] information, which is essential details [2:42] and medical contacts for quick [2:43] reference. Number two, what to do next, [2:46] which is the top three actions to pursue [2:49] or the top three questions to ask your [2:50] doctor. [2:52] Number three, what we know, which is the [2:54] key facts about the diagnosis and [2:56] treatment for the patient. [2:59] And then there's an explanation of [3:00] medical terms in plain English, and a [3:03] care log, which is compact dated history [3:06] of updates as you find out more [3:08] information. [3:10] Okay, so let me show you exactly how to [3:12] install the skill and get this kind of [3:14] brief in a demo [snorts] account. [3:17] All right, so this is a demo ChatGPT [3:19] thread. [3:20] And all I said was install the skill at [3:23] github.com/petergyang/fcancer. [3:27] And I'll link the skill description in [3:29] the video as well. [3:31] After installing the skill, you can [3:32] start something like this. F Cancer, my [3:35] mom was just diagnosed with stage three [3:36] breast cancer. [3:38] Here are her markers, ER and PR [3:40] positive, and so on and so forth. [3:43] And here's her oncologist. And then it's [3:45] going to go ahead and create the family [3:47] brief. [3:48] So, let's take a look at the brief right [3:49] now. [3:50] So, here's a hypothetical example of the [3:53] brief. [3:54] And again, it starts with the patient [3:55] information, who the patient is, how old [3:57] they are, what their phone number is, [3:59] who is on their current care team, and [4:02] some key phone numbers and emails that [4:03] you can reach out to. [4:05] And also, if you're getting a second [4:07] opinion, who that person is. [4:09] And then it goes into what to do next. [4:11] So, questions you can ask her doctor. [4:13] And what we know, which is the [4:15] diagnosis, the biomarkers that the [4:17] cancer has, the stage of the cancer, [4:20] treatment status, and so on and so [4:22] forth. [4:23] And any kind of medical terms that it [4:24] has, it tries to explain down here, [4:26] right? And finally, there is a care log. [4:29] So, again, having this simple Google [4:31] document that you can update and share [4:34] with the patient and other caregivers, I [4:36] just found could be extremely [4:37] invaluable. [4:38] In helping me to make sure that I ask [4:40] the right questions to my mom's doctors [4:42] in the 15-minute or 20-minute [4:44] appointments that we have with them. [4:46] Going back to the ChatGPT conversation, [4:49] once you have this brief setup, you can [4:51] say something like our appointment with [4:52] the doctor is coming up. Based on our [4:54] brief, what are the three most important [4:56] questions that we should ask? And here's [4:58] the three questions over here. [5:00] So, again, set up the brief first. It [5:02] doesn't have to have all the [5:03] information, but have this one-page [5:06] source of truth that can get updated as [5:09] the patient goes through diagnosis and [5:11] treatment. [5:12] Okay, so after you set up the brief, [5:14] step two is to organize the patient's [5:17] medical records. [5:19] Your brief is really only as good as the [5:21] data behind it. So, the more records you [5:23] have on the patient, the more [5:25] personalized and up-to-date AI's [5:28] responses can be. [5:30] Unfortunately, my mom has had cancer [5:32] three times before. So, her history was [5:34] scattered across different hospitals and [5:36] portals. [5:38] And, you know, I wish that we requested [5:40] all three types of records right away [5:43] instead of waiting. [5:44] So, three types of records are written [5:45] records, like previous scans and [5:48] pathology reports, and treatment [5:49] history. [5:50] Imaging files, like CT, PET, and other [5:53] scans. And this stuff is usually [5:55] delivered in a format called DICOM. [5:57] And it takes about like two or three [5:59] weeks to get this stuff. And they're [6:00] usually hand-delivered to you by the [6:02] hospital in a thumb drive or even a CD [6:05] ROM. So, start the request process [6:07] early. [6:08] And last but not least, the pathology [6:10] blocks or slides of the actual tumor [6:13] from the biopsy. You know, this is the [6:15] most precious material for sequencing [6:18] the tumor, figuring out exactly what it [6:20] is, and also for getting second [6:21] opinions. And this also takes the [6:23] longest to get. [6:25] So, if the patient has had cancer before [6:27] like my mom, [6:28] request these as early as possible [6:30] because they have to potentially dig it [6:32] up in old warehouses to find them. [6:35] My mom, for example, is still waiting [6:37] for the pathology block from a prior [6:40] cancer. [6:41] And what I do is I keep all my mom's [6:43] medical records and scans in a shared [6:45] Google Drive that looks something like [6:47] this. [6:49] And I can get ChatGPT by the way to help [6:51] organize all of this stuff. [6:53] But basically, pointing ChatGPT and [6:54] pointing AI at this Google Drive folder [6:57] and my mom's brief has been critical in [7:00] grounding its answers on her actual [7:02] history instead of on generic medical [7:05] advice. [7:06] So, again, you can go back into ChatGPT [7:09] and you can just say, "We've started [7:11] collecting the patient's medical [7:12] records. They're spread across two [7:14] hospitals. What should we request first [7:17] and how long will each take?" And then [7:18] ChatGPT will give this pretty awesome [7:19] table of where you can get stuff. [7:22] And if you see something like, "Hey, [7:23] we're going to request these records [7:24] from the specific hospitals like UCSF [7:26] and Stanford," [7:28] then [7:29] it will actually provide the links for [7:31] you to request the records and maybe [7:33] even send emails for you, right? Taking [7:36] care of all of the paperwork for medical [7:38] diagnosis. [7:40] All right, let's keep going. So, after [7:41] you get the medical records, step three [7:44] is to research treatment options. [7:46] And my mom's diagnosis with cancer [7:49] follows a pretty typical patient's [7:51] diagnosis. [7:52] So, first, her doctor saw a suspicious [7:54] tumor [7:56] in a CT scan and then more follow-up [7:58] scans. [7:59] Then they did a biopsy of the tumor, [8:01] which confirmed that it was cancerous. [8:04] Then they had a tumor board meeting, [8:06] which is basically like a Zoom call [8:08] where her oncologist, radiologist, [8:10] surgeon, and other specialists discussed [8:13] her case and what treatment options [8:15] there are. [8:16] And usually treatment starts with [8:18] standard care based on guidelines like [8:21] NCCN in the US, [8:24] which typically is some combination of [8:26] chemotherapy, radiation, surgery, or [8:29] immunotherapy. [8:31] Now, unfortunately, it's also typical [8:33] for this process to take months. In [8:36] fact, it's been 2 months since my mom's [8:38] first CT scan, [8:40] and she's still waiting to find out [8:42] whether this cancer is a reoccurrence of [8:44] her breast cancer or something else. [8:47] Now, I have a lot of respect for doctors [8:50] and have many friends who are physicians [8:52] and nurses. [8:53] But, the system is just very, very [8:55] difficult. Working with my mom's medical [8:58] team has sometimes felt like working [9:00] with a co-worker that doesn't respond to [9:02] emails or calls [9:04] and only communicates through voicemails [9:07] and faxes. [9:08] In fact, I was pretty frustrated the [9:09] other day and I tweeted this. You know, [9:12] there's all kinds of crazy breakthroughs [9:15] and people tweeting that AI has cured [9:18] cancer on X, which by the way is total [9:22] And in the meantime, you know, what I'm [9:24] doing is [9:26] faxing documents and hand delivering [9:29] images on CD-ROMs. So, there's a massive [9:32] disconnect between what state of the art [9:34] treatment looks like and what standard [9:37] care and procedure looks like. And as a [9:39] caregiver, I think it's my job to really [9:42] fight for my mom and try to get her the [9:44] best care possible and navigate a lot of [9:46] the bureaucracy. [9:48] So, as you wait for the doctors to get [9:50] back to you, you can feel pretty [9:52] helpless because there's a lot of [9:53] uncertainty and ambiguity, and maybe the [9:56] tumor is growing this whole time. But, [9:58] there's actually stuff that you can do [9:59] in the long wait. [10:01] And specifically, I think there's like [10:02] three steps. [10:04] So, step one is to understand standard [10:07] care and try to get a second opinion. [10:10] You know, in the US, doctors typically [10:12] follow NCCN guidelines for first-line [10:15] cancer treatment. And you can read them [10:17] yourself with a free account. [10:20] So, once you have a diagnosis, [10:22] ask AI to pull standard care options [10:25] from public sources, [10:27] and basically try to find out what the [10:29] standard treatment is. [10:30] So, here's a quick example. You can [10:31] basically just ask AI, ask My Cancer [10:33] Coach, [10:34] "What does standard treatment look like [10:35] for a case like the patient's? [10:38] And where do we get a second opinion?" [10:40] And it can give you basically, you know, [10:42] usually surgical treatment, often [10:43] including surgery, [10:45] based on her specific markers. [10:47] And then there's also hormone blocking [10:49] treatment because uh this hypothetical [10:51] example, the breast cancer is ER and PR [10:53] positive. [10:55] Okay, but the point is to kind of [10:57] research standard care, get second [10:59] opinion options, so that you can be [11:01] informed, and basically, you know, have [11:04] evidence-backed reasoning around the [11:06] type of treatment that the patient [11:07] should get. [11:09] And for second opinions, I recommend [11:11] getting it at a major cancer center like [11:13] Anderson in Texas, MSK in New York, [11:16] Dana-Farber in Boston, or Stanford and [11:19] UCSF in the Bay Area. [11:21] By the way, getting a second opinion is [11:23] another long and involved process. It [11:26] can take up to a month. [11:28] And getting all the medical records and [11:30] images organized in step two is [11:32] extremely helpful for the process to get [11:34] a second opinion. [11:35] So, right now we're right in the middle [11:37] of getting a second opinion from UCSF, [11:39] for example, and I've been using AI to [11:42] basically organize all of her medical [11:43] records, to send emails, [11:46] draft cover letters explaining her case. [11:49] Okay, so another step that you can take [11:51] is to get the tumor sequenced. [11:54] Cancer is actually more than 100 [11:56] different types of diseases, and each [11:59] person's tumor can behave differently. [12:02] That's why I think it's really important [12:03] to get a tumor sequenced, to look for [12:06] DNA, RNA, and other markers that can [12:09] point to better treatment. [12:11] So, right now we're planning to use this [12:12] company called Boston Gene [12:15] uh that other people have recommended. [12:17] It's going to cost us around $4,000 to [12:19] do this, but basically we're going to [12:21] get back a report that looks something [12:23] like this. [12:24] Let me zoom in a little bit. [12:26] And with a bunch of biomarkers, which [12:28] could hopefully inform treatment options [12:31] down the line. [12:33] Okay? And you can also go ahead and go [12:35] in ChatGPT and ask AI where tumor [12:39] sequencing could be useful right now, [12:40] and also help me find top clinical trial [12:43] options. [12:44] Which is the next option here. [12:47] So, again, in the beginning when your [12:49] patient is diagnosed, [12:51] the doctors are going to want to pursue [12:52] standard care options because they're [12:54] proven, right? But some of the most [12:56] innovative treatments out there are [12:57] still going through clinical trials. And [13:00] by the way, you should be very careful [13:01] about clinical trials because [13:03] some stage one and two trials are [13:04] actually unproven and could actually [13:06] make things worse. [13:08] But the point is you want to get AI and [13:11] you want to use your own skills to [13:13] research the clinical trial options that [13:16] are available, right? So, going back to [13:19] our conversation here, I asked find the [13:21] top three clinical trial options in the [13:22] US or Canada. [13:24] And here it is, the top three options. [13:27] And you should obviously read through [13:28] the options. [13:30] And potentially some of this stuff is [13:32] actually near [13:33] where the patient lives [13:35] and is also relevant to the patient's [13:37] disease. [13:38] Again, because AI has full context of [13:40] the brief and the medical records, [13:42] it can actually recommend useful [13:44] clinical trials. [13:46] So, what my skill does is it points to [13:50] clinical [13:51] trials.gov, [13:54] which has a database of all US-based [13:57] clinical trials that you can search for [14:00] or get AI to search for. [14:02] Okay, again, while you wait for the [14:04] diagnosis or even during treatment, it's [14:06] always useful to understand standard [14:08] care, get a second opinion, get a tumor [14:11] sequenced, and also research clinical [14:14] trials. Because all this stuff takes [14:15] longer than you think it does. So, it's [14:17] good to get started early. [14:19] All right, step four is to find experts [14:22] to have in your corner. [14:24] You know, one of the best things that [14:25] come from sharing my mom's story online [14:28] has been the people who have offered to [14:30] help. I think at least a dozen people [14:33] have emailed me with their stories and [14:35] offered to help. And I want to give a [14:36] special shout out to three people. So, [14:38] Ari from Open Cancer who made [14:40] introductions to people who could help [14:42] us navigate testing and pathology [14:44] review. [14:45] Bailey from the Sege Foundation who [14:48] helped us prepare questions about the [14:50] diagnosis, tissue testing, and treatment [14:52] options. And even real oncologists like [14:55] Dr. Odabashian [14:57] and other folks [14:58] who emailed me and DM'd me on Twitter to [15:01] provide their perspectives on my mom's [15:03] case. [15:04] So, to keep them in the loop, I [15:05] basically use ChatGPT to draft short [15:08] updates [15:09] like this. [15:10] So, I say, "Hey, a French dude says to [15:12] oncologist, [15:13] can you draft a short email update?" [15:15] And here is the email update with the [15:17] key facts. [15:18] And I BCC basically my board of experts [15:21] and they offer their opinions and [15:23] advice. [15:24] You know, most families who get cancer [15:27] like to keep it to themselves. But, what [15:28] I found is that sharing the story and [15:31] sharing the journey online [15:32] or even just with friends [15:34] really empowers other people to come and [15:36] help. [15:37] Cuz a lot of people are actually going [15:38] through this disease or have gone [15:40] through it before. [15:42] All right, the last step is to protect [15:44] quality time with your family. You know, [15:46] when I found out about my mom's cancer, [15:48] I took a week to stay with her. We had [15:50] many walks, great meals, and chats that [15:53] I'll cherish forever. [15:55] And I also call my parents every day so [15:56] that they can talk to the kids and feel [15:58] happy. [15:59] Now, if you're in your early 40s like [16:01] me, you probably have young kids and [16:04] aging parents, too. [16:05] Right when it's easiest to get caught up [16:07] chasing more money and titles and [16:10] focusing on your job and career. [16:12] But, at some point in your life, you [16:13] know, especially if you're successful, [16:16] recognize that your time becomes more [16:18] valuable than money. [16:20] And time with your loved ones is [16:22] actually easiest to neglect and postpone [16:25] because there's really a deadline tied [16:27] to it and you can't actually quantify it [16:29] like you can quantify how much money [16:30] you're making. [16:32] But basically going through this journey [16:33] with my parents [16:34] has made me realize just how precious [16:36] time is. [16:37] There is this great comic strip that [16:39] illustrates this, right? It shows a dad [16:42] growing up and taking care of his [16:45] daughter. [16:46] And you know, the dad's getting older, [16:47] the daughter's getting bigger. [16:49] And before you know it, the kid has [16:51] moved out and you're getting old. [16:53] So, my girls are probably, I think, [16:56] around this age and maybe this age right [16:59] now. [17:00] And it's only a matter of time before [17:01] they get bigger and they leave the [17:03] house. I I think if there's one thing [17:04] that I made you realize in this video [17:07] is to recognize that time with your [17:10] spouse, your kids, and your parents [17:12] is very precious and limited and can [17:15] pass by before you know it, right? So, [17:17] definitely cherish those moments. [17:19] And I have to remind myself about this, [17:21] too. [17:22] All right. So, you can find my F Cancer [17:24] skill in the video description. [17:26] And if you're going through cancer or [17:27] know someone who is, please know that [17:29] you're not alone. And as much as you [17:31] feel comfortable, share your journey [17:33] with others and you'll be surprised at [17:35] how much they're willing to help. [17:37] This isn't a typical video that I'm [17:38] making, but I hope you enjoyed it [17:40] regardless. [17:41] And if you want more personal videos [17:43] like this, [17:44] please like and subscribe [17:46] and let me know in the comments. [17:47] I'll see you next time.