XLABS By Travis Dirks PHD, CTO, XLabs AI March 26, 2020 - - PowerPoint PPT Presentation

Covid-19: How To Rapidly Get To The Only Possible End Game – Herd Immunity

Three new game-theoretic strategies that improve on ‘Sheltering in Place’ to battle the current Covid global

• pandemic. All three strategies balance the impact to both human lives and the economy within weeks.

1

By Travis Dirks PHD, CTO, XLabs AI March 26, 2020 Contributors

• Dr. Lorna Tumwebaze*, MBChB, MD/MPH, DPH, Epidemiologist & Global Policy Advocacy, The Gates Foundation
• Dr. Monica Bucci*, MD & Director of Clinical Research at The Center for Youth Wellness

Emmy Sobieski, CFA, MBA, 25-year institutional investor focused on disruptive technologies Radhika Dirks, PhD, Top 30 Woman in Advanced AI & CEO XLabs AI

XLABS

• Herd immunity is the only destination in sight for Covid-19. Ifherd immunity is not possible, there is no end game. Everything else is just living with the problem. It is very important to

understand that there is literally no other way to win.1

• Herd immunity is defined as:
• 1. a significant number of people have had the disease and can no longer pass it on, and
• 2. the chances of someone infected with the virus meeting someone, who has neither been infected nor is mortally vulnerable, is near zero.
• Vaccines are a shortcut to herd immunity. But ETA to Vaccines is: ~12 months.

There Are 3 Other Tech-enabled Strategies That Can Be Ready To Implement In 2-3 Weeks That Save Lives & Preserve The Economy

We Need To Get To Herd Immunity ASAP

2

XL XLab abs

¹Fundamentally we can not control whether the end game of herd immunity is available to us. Nature has decided already, and we will learn as time passes. What we can control is which potential road to herd immunity we are on and what death & economic toll we will pay.

Solutions: Fast Track to Herd Immunity

Problem with Current Strategy:

• Basic Social Distancing is basically Freeze Tag. Everyone is

frozen.

• It will work, but experts state that it needs to be in effect for 12

months: unviable for the economy

• South Korea’s ‘Test Broadly and Quarantine’ saving the economy

from the worst but creates a near permanent state of disease & lockdowns

• What we must do now is clear and MUCH easier technologically

than creating a vaccine. The RouteForward:

1 .

3

XL XLab abs

1. Invest heavily in vaccines and in raising the thresholds at which our hospitals breakdown & mortality rates jumps from <1% to 5-10%. 2. Invest aggressively in the technology & policy needed for Social distancing 2.0, 3.0, Change the Odds

SEVERE (potential depression) SHORT (weeks) NOT LONG (fewmonths) VERY LONG (could be years) MEDIUM (minordip) MINOR (very littlechange)

ECONOMIC IMPACT TIME IN CRISIS

Social Distancing 3.0 (Improved Strategy 3 Tech Enabled) Social Distancing 2.0 (Improved Strategy 1; Tech Enabled) Change the Odds (Improved Strategy 2 Tech & Science Enabled) Social Distancing (Current in US, UK, etc.) Test Broadly & Quarantine (Current in South Korea)

Road 1

Hospital Overwhelm: Mortality Rate ~5-10%.If herd immunity happens at around 90%

• f the population recovered, then this route equates to 15,000,000-30,000,000

deaths in the United States, mainly among those over 70 year s old. This is not a road we want to be on.

Road 2

Hospitals Functional: Mortality Rate ~0.5-1%.Ifherd Immunity happens around 90%, this route equates to 1,500,000 to 3,000,000 deaths in the United States, again mainly among those over 70 years old. Until a vaccine or other cure is available, this is the best road we have in our power to take.

Road 3

Vaccine or other Cure: Mortality Rate ~0.0001%. can’t get to without heavy investment and passage of time for testing. Ifherd Immunity happens around 90%, this route equates to 230 deaths in the United States. Age distribution unknown. This is the road we would dearly love to be on, but won’t be available for 12 months.

There are only three different mortality rates available to us: roads to herd immunity. We need to be on Roads 2 & 3. .

There are Only 3 Ways to Get To Herd Immunity

4

XL XLab abs

• 5

10 15 20 25 30 35 Road 3: Vaccine or Other Cure Road 2: Hospitals Functional Road 1: Hospitals Overwhelmed

Deaths (Millions)

Deaths to Reach Herd Immunity @ 90% Infected Rate in United States

33 Million 3.3 Million 330

Covid-19 The Complete Playbook – 3 Unacceptable & Unavailable Strategies

• 1. Do Nothing - Let nature takeits course

NO Vaccine, No Flattening the curve. This potentially leads to 15-30M

• dead. Not an option.
• 2. Develop aVaccine

People have been working on a coronavirus vaccine for YEARS. It will not come tomorrow. Ifwe discovered the vaccine today, estimates are it would take 18months to get it into enough people’s hands to help. While this is the best strategy and should be heavily invested in, it must be paired with one of the strategies for flattening the curve until the vaccine is available.

• 3. Raise theThreshold

Raise the Threshold at which hospitalization service breaks down to the Hospital Overwhelm rate of death, i.e., raise total nationwide hospital treatment capacity. Downside –huge lag time of months until supply chains adjust to the critical equipment needs Like ventilators and ICU beds, this is necessary but is not an immediate alternative.

TIME SINCE FIRST INFECTION

DO NOTHING

NUMBER INFECTED ECONOMY (GDP)

5-10% of known infections dead. Death toll in the millions. Length of Crisis (Months) Known cases above threshold - Mortality 5-10% Known cases below threshold - Mortality 0.5-1% 5

XL XLab abs

Hospital Breakdown Threshold Total Real Infections Total Seeking Treatment

These 3 strategies are all unacceptable or currently unavailable

Flatten theCurve Flatten the Curve: Find a way to slow the spread or otherwise keep the number of critical hospitalizations below threshold.

• 4. Basic Social Distancing

Very Negative Economic Impact and Prolongs the Emergency. Social distancing is a game of freeze-tag, with everyone frozen. It will work, it will also wreck the economy and cause incredible negative downstream effects.

• 5. Test Broadly and Quarantine

Lowers Economic Impact, but still Prolongs the Emergency. The South Korean Strategy – test broadly and often. Pull the infected quickly into

• quarantine. Everyone else goes about life a little more carefully. Has the

great advantage of not decimating the economy. The biggest downside is that herd immunity is never reached. You have only paused the game. T I M E S I N C E F I R S T I N F E C T I O N R E A L N U M B E R I N F E C T E D ( N U M B E R W EK N O WA B O U T ) ECON OMY

0.5-1% of known infections dead. Death toll in the tens of thousands. Length of Crisis - New Normal Known cases above threshold - Mortality 5-10%

SOCIAL DISTANCING AND QUARANTINING

Known cases below threshold - Mortality 0.5-1% 6

XL XLab abs

TIME SINCE FIRST INFECTION NUMBER INFECTED ECONOMY (GDP)

Hospital Breakdown Threshold Basic Social Distancing Test Broadly and Quarantine

Both these strategies have such broad economic impact because they manipulate the entire society to indirectly effect the total real infections. But their ultimate goal is really to reduce the total seeking treatment. What if we could manipulate the curve that must stay below threshold directly?

Current Covid-19 Playbook – 3 Working But Painful Strategies

Total Real Infections Total Seeking Treatment

T I M E S I N C E F I R S TI N F E C T I O N R E A L N U M B E R I N F E C T E D ( N U M B E R W EK N O WA B O U T ) ECON OMY

Social Distancing 2.0 - Length of Crisis (Months) Social Distancing 3.0 - Length of Crisis (Weeks) Known cases above threshold - Mortality 5-10% Known cases below threshold - Mortality 0.5-1%

SOCIAL DISTANCING 2.0 + 3.0

0.5-1% of known infections dead. Death toll in the tens of thousands.

• 8. Change the Odds

While a bit of a moonshot, it may be possible to isolate and understand Covid-19’s immune system response in the asymptomatic and immune young population, with the goal of replicating that response into at-risk

• populations. Intentionally changing the odds of the disease response

would very significantly drop the mortality rate and narrow the curve while shrinking the length and magnitude of the economic impact. This has a similar lead time to a vaccine.

• 7. Social Distancing 3.0

Super Fast Track to Herd Immunity: Lowers Economic Impact and Dramatically Shortens the Emergency: Similar to chicken pox parties, if we can identify those who are highly invulnerable to the Coronavirus, we can do the same and dramatically speed the route to herd immunity. If we can identify those who will not need hospitalization with high precision, then those people can intentionally & voluntarily self-infect to drastically speed herd immunity. While this may sound Machiavellian, many young people are already doing this, albeit without the important science and information behind who should, who shouldn’t and how it should be done to protect lives.

3.0

7

XL XLab abs

• 6. Social Distancing 2.0

The Fast Track to Herd Immunity: Lowers the Economic Impact And Shortens the Emergency. Shelter in Place only for those who are high risk and let everyone else go about life to reach herd immunity fast. This can be accomplished only if with high precision, we can pick out those infected who will need hospitalization. Those people must self- isolate along with those who care for or are in contact with the vulnerable.

2.0

TIME SINCE FIRST INFECTION NUMBER INFECTED ECONOMY (GDP)

Hospital Breakdown Threshold Total Real Infections Total Seeking Treatment Social Distancing 2.0 Social Distancing 3.0

New Covid-19 Playbook: 3 Improved Strategies to Save Lives & Economy

Artificial Intelligence (AI) is ideally suited for a rapid response for allowing the three ideal strategies to solve Covid -19, in the short term. As an example, XLabs’ infrastructure, XCore, is built for solving complex impactful problems using real -world data. XCore’s AI is a self-learning AI engine designed to automate scientific breakthroughs and is already helping solve some of the world’s most important and impa ctful problems, such as cancer and drug discovery. Our AI engine has been used to accelerate new cures for cancer (years to weeks), as well a s develop novel measures for tracking and tackling past global epidemics like Ebola. Here is how XLABS AI combined with broad and ubiquitous panomics measurement can deliver years of scientific progress in a matter of weeks and enable the three best Covid-19 strategies: Social Distancing 2.0, 3.0 and Change the Odds

Covid-19: In Weeks, AI can help implement the path to the 3 Improved Strategies

Technological Roadmap: How AI can help with Immediate Solutions

Because XLabs AI Infrastructure, XCore, is built for solving complex impactful problems using real world data, it is ideally suited to solve the technological challenges needed to enable all of these strategies. Send Safe Patients Home: Impactful. Timeframe: 2weeks. XCore is capable of creating a high-precision AI that can predict who will recover with minimal care, so the patients can be sent home safely. This technology will not solve the shortage of ventilators and other key ICU equipment, but it will relieve significant pressure on hospital staff, free up hospital beds, and increase hospitals’ ability to deal with the current crisis. . Social Distancing 2.0: Very Impactful. Timeframe: 4-6 weeks. Risk stratification 1 Social Distancing 2.0 can deliver a large improvement in lives lost and can dramatically decrease the negative economic impact of responding to the

• Coronavirus. XCore is capable of creating a high-precision AI that can separate

at-risk demographics into those who, if infected, have a very high chance of needing intensive care, and those who are actually low-risk for needing the ICU. This technology, combined with quarantining of the infected, would allow a form

• f selective social distancing -- in which the majority of the population goes

about life as usual, reaching Herd Immunity in months, while those who are highly at risk, or care for the highly at risk, Shelter-in-Place and carefully avoid

• utside contact. For example, the mortality rate of 80+-year-olds is ~14%,

meaning that if 1,000 people are infected, 140 will die. And 300-500 will need critical hospital care. Many will recover on their own from the illness. What if we could identify these groups ahead of time? The time to start using this technology is mainly determined by the will to collect the needed data. .

• 2. 0

9

XL XLab abs

Social Distancing3.0: Very Impactful. Timeframe:6-10weeks. Similarly, those below the age of 40 have a mortality rate of 0.2%. This means that if 1,000 are infected, 2 will die, perhaps another 5-10 will need critical hospital care. What if we could Identify those 10 who are at risk, and ask only them to shelter- in-place? This would enable Social Distancing 3.0, in which aggressive quarantining is no longer necessary, and Herd Immunity can be reached very quickly. This is because the risk for those not required to Shelter-in-Place is now below that of the seasonal flu. From XCore’s perspective, this is a similar problem to that required for Social Distancing 2.0. It does, however , require 10x more data to be collected, since the adverse events are so rare. . Change the Odds: Impactful. Time Frame: months Change the Odds by using XCore’s AI to reveal high-priority targets for medical

• intervention. The above-mentioned technologies to achieve Social Distancing

2.0 and 3.0, however, do not require any medical treatment to be developed, Change the Odds does require the development of new targeted medical treatments that use AI. Simply put, the concept is to choose a cohort of 100 low-risk and 100 high-risk individuals who are recently infected and observe the progression of the disease with broad panomic measurements. Then we can use the dataset to identify the differences and similarities between the low- risk, the risk that recovered well, and the high- risk that needed ICU care. Similar to our work with cancer, these differences and similarities reveal high- priority targets for medical intervention. This could enable an at-risk patient's immune system to respond more like that of a low-risk patient. Due to the targeted nature of the development of this targeted medical treatment, the timeframe for reaching a solution is likely significantly shorter than a generalized vaccine development program. .

3.0

1

XL XLab abs

None of the above is magic. It will require a concerted effort to broadly collect the data needed to make it possible. With investment and will it can be done on a scale of weeks.

Technological Roadmap: How AI can help with Immediate Solutions

w w w . x l a b s . a i Contact:

• Dr. Travis Dirks, CTO XLabs AI

Travis@XLabs.ai

XLABS

ACT NOW

• USE THE RECOMMENDA

TIONS HERE TO SA VE LIVES & ECONOMY NOW

• HELP OUR EFFORT TO IMPLEMENT THESE SOLUTIONS WITH:

DA TA COLLETION PARTNERSHIPS CAPITAL DEPLOYMENT PARTNERS

• SHARE WIDELY