Tesla (2Q25 Minutes): Can the Grand AI Vision Prop Up a Faltering Core Auto Business?
The following is$Tesla(TSLA.US) the Tesla FY25Q2 earnings call minutes. For earnings interpretation, please refer to "Tesla's 'Life-or-Death Question': With AI's Starry Sea, Does Car Manufacturing Still Matter?"
1. Review of Core Financial Information
Dolphin's Take on Tesla 2Q25 Earnings Call: Fundamentals Still Weak in the Near Term, but Steady Progress on Moonshot Projects like Optimus and Robotaxi/FSD：
1. Auto Business: Near-Term Outlook Remains Weak：
a. Mass production of the lower-cost Model 2.5 has been delayed.
b. The phaseout of the $7,500 IRA subsidy in the U.S. is expected to weigh on demand over the coming quarters.
c. Tariff-related cost pressures will be fully reflected in future quarters due to a lag in revenue accounting, likely pushing short-term costs higher.
d. Regulatory changes to emissions credits are set to drive a decline in both Q2 and forward-looking revenue projections.
2. Energy Business：
AI Demand Offers Long-Term Promise, but Short-Term Challenges Persist
Tesla expects AI and data center-related demand to boost its energy storage business over time. However, near-term growth remains constrained by the impact of the Inflation Reduction Act (IRA), especially in the residential segment. The zero-carbon incentive program is now scheduled to end earlier than expected—by year-end.
In addition, energy storage costs are also affected by tariffs, although Tesla has announced plans to localize LFP battery production in the U.S. to mitigate this.
Despite still being in the planning phase, Tesla's long-term visionary projects — including Optimus and Robotaxi/FSD — have made encouraging progress：
3. Optimus: Progress on Track
Next 3 Months: Tesla plans to unveil the Optimus Gen 3 prototype within the next three months and initiate small-scale production.
2026 Outlook: Mass production is expected to begin early next year, but initial financial impact will be limited as costs will still outweigh revenues. Many technical and process challenges remain.
Long Term: Tesla aims to ramp up to a monthly production capacity of 100,000 units within five years.
Management cautioned that since Optimus is a ground-up product, production scale-up will be constrained by the weakest links in the supply chain and internal processes. Early-stage manufacturing is unlikely to contribute meaningful revenue and will primarily serve to identify and resolve issues.
4. Robotaxi: Key Milestones and Ambitious Expansion Plan
① Current Progress: This quarter marked a major milestone as Tesla launched a paid Robotaxi service in Austin with no safety driver. The operating area has expanded and will continue to grow in the coming weeks—eventually outpacing existing competitors in geographic reach.
② By Year-End: Tesla is actively working with regulators in the Bay Area, Nevada, Arizona, and Florida. Upon approval and confirmation of safety benchmarks, the company aims to roll out its Robotaxi service to most of the U.S., with the goal of covering more than half the U.S. population by year-end.
③ Financial Impact: The Robotaxi business is expected to begin generating material financial returns by late next year.
④ Scaling Strategy: Tesla will initially use its balance sheet to fund fleet expansion. Once revenue reaches scale, it plans to tap into external financing such as debt.
5. FSD (Full Self-Driving): Progress on Both Software and Hardware
① Software: Tesla continues to make strong advances in FSD software. The plan is to expand the model’s parameter size by roughly 10x—an engineering challenge, especially under memory bandwidth constraints. However, the company believes this is achievable on existing HW4.0 systems.
② Hardware: The AI 5.0 chip is expected to enter mass production by the end of next year and holds significant potential. Tesla is also working on integrating Dojo 3 with the upcoming AI 6 inference chip, creating a unified architecture that could power both vehicles and Optimus robots—or be scaled up in clusters for data center use.II. Detailed Content of the Earnings Call
2.1 Key Information from Executive Statements
I. Musk's Speech:
1. Car Sales:
Notably, the Model Y became the best-selling model in Turkey, the Netherlands, Switzerland, and Austria in June this year, and continues to hold the position of the best-selling car globally. The autonomous driving feature is a key selling point.
2. Robotaxi:
① Current Progress: This quarter has been a milestone for us. We successfully launched the Robotaxi service in Austin, achieving driverless rides for paying users. This service has now expanded its operational area and is expected to further expand in the coming weeks, covering areas far beyond existing competitors.
② Future Plans:We are actively advancing regulatory approvals in the San Francisco Bay Area, Nevada, Arizona, and Florida. Once permitted and safety requirements are met, we will gradually roll out autonomous travel services across most of the United States. Our goal is for the Robotaxi service to cover more than half of the U.S. population by the end of this year.
All of this is contingent on regulatory approval—we are technically ready. Meanwhile, we will proceed cautiously. Although the number of service vehicles and regions will expand rapidly, we will ensure that safety is not compromised.
3. FSD:
Regional Progress:
① Europe: Although we have not yet obtained supervised FSD (Full Self-Driving) certification in Europe, we believe that once approved, sales in the European market will increase significantly. We are currently working with Dutch regulators and are close to approval, after which we will need to submit to the EU for review, a rather complex process.
② China: We face a similar situation in China, where we are currently unable to offer FSD functionality, but we are working to overcome regulatory hurdles. FSD is one of the largest new car demand drivers globally.
③ United States: In the U.S., as we gain more confidence in safety across different regions, the regulatory requirements for driver attention on the road will gradually relax. Currently, some users temporarily disable Autopilot to bypass system warnings and then restart it after completing tasks, which actually reduces safety. In the coming weeks, the FSD experience will see significant improvements, especially for users outside of Austin. AP features will gradually integrate into the user experience:
We are currently focused on deploying the Austin Robotaxi, which has caused some delays in the official version of Autopilot in other regions, but now we will gradually integrate the premium features from the Austin Robotaxi into the overall user experience.
Overall, everything revolves around the main theme of "autonomous driving": hardware is the foundation, but autonomous driving is the value amplifier. Additionally,
FSD Software Iteration:
In terms of FSD, we continue to make significant progress in software. We plan to expand the model's parameter scale by about 10 times—a technical challenge, especially when limited by memory bandwidth. However, we believe we can achieve this goal on existing hardware.
4. Energy Business: Despite Policy Headwinds, Demand Remains Strong with High Long-term Certainty
a. Megapack (large-scale energy storage) capacity is rapidly expanding, with performance upgrades, and Q2 deployment reached a record high, offsetting tariff and supply chain pressures.
b. The scale of battery demand is enormous, and pairing U.S. grids with batteries can more than double annual power generation, with high long-term growth certainty (for example, the continuous power demand of the U.S. grid is about 1 terawatt, while average electricity use is less than 0.5 terawatts. With energy storage, power plants can operate at full capacity 24 hours a day, doubling energy output, which is an extremely important potential).
The energy business also performed very strongly. Despite facing tariff and supply chain pressures, Megapack capacity is rapidly expanding, and the next-generation product's performance will further improve, with second-quarter deployments reaching a new high. The scale of battery demand far exceeds people's imagination. For example, the continuous power demand of the U.S. grid is about 1 terawatt, while average electricity use is less than 0.5 terawatts. With energy storage, power plants can operate at full capacity 24 hours a day, doubling energy output, which is an extremely important potential.
5. Optimus:
Optimus Iteration:
Regarding the humanoid robot Optimus, we are currently at version 2.5, moving towards Optimus 3, which we consider a truly mature design. We plan to complete the Optimus 3 prototype by the end of the year and begin mass production next year, aiming for an annual production of 1 million units within the next five years.
This product presents extremely high technical challenges, with each component needing to be independently developed from physical principles, including motors, gearboxes, sensors, and neural network training. We will apply the AI inference optimization technology accumulated in automotive autonomous driving to humanoid robots.
I want to emphasize that Tesla is the world's leading "real-world AI" company. Compared to competitors like Waymo, although the latter deploys a large number of sensors, their effectiveness is far inferior to ours.
We are far ahead in AI inference efficiency. I believe the core metric for measuring AI should be "intelligence density per unit memory bandwidth," rather than simply the number of parameters. In this regard, Tesla leads the world.
Additionally, I am aware of the situation with xAI's product Grok. Grok 4 is a very powerful model, but its operating costs are extremely high. Tesla's advantage lies in "intelligence density"—this will become a decisive factor in the future development of AI. In summary, 2025 has been a very significant year so far. We have achieved many autonomous driving breakthroughs that were once doubted by outsiders. Although not always on time, we deliver on our promises. The Tesla team is fulfilling the future, and I firmly believe we are poised to become the most valuable company in the world.
II. CFO's Speech
1. Second Quarter Performance:
① In terms of automotive revenue:
Despite a decrease in regulatory credit revenue, total automotive revenue increased by 19% quarter-over-quarter, while total deliveries only grew by 14%. This was mainly due to the average selling price (ASP) increase brought by the new Model Y. With product mix optimization and improved fixed cost allocation, despite increased tariff costs, profit margins also improved quarter-over-quarter.
② Second Quarter Tariff Impact:
We have already seen the impact of tariffs on the income statement. Quarter-over-quarter, tariff costs increased by about $300 million, with approximately two-thirds affecting the automotive business and the rest impacting the energy business. However, due to the lag in sales accounting, the full impact will be seen in the coming quarters, so short-term costs will rise. Although we are doing our best to mitigate these impacts, the tariff environment remains uncertain.
③ Energy Business:
The profit margin of energy production improved quarter-over-quarter, despite a decrease in deployment volume, mainly due to the continued advancement of high-margin projects. We achieved the highest gross profit to date for this business. It should be noted that overall deployment volume will continue to show quarterly fluctuations.
As Elon mentioned, industrial storage will play a significant role in the growth of AI and data centers. As AI applications develop, energy demand is rapidly increasing because AI applications have a large energy demand. AI and data centers have a surge in energy demand, and customers have realized the importance of energy storage. Despite this business being most affected by tariffs, customers still show strong purchasing intent.
However, the "Big Act" also poses challenges to the energy business, especially residential energy storage, as the zero-carbon incentive policy will end early by the end of the year.
④ Service Business:
The gross profit of service and other businesses improved, mainly due to the performance of the supercharging network, insurance, and after-sales centers.
⑤ Operating Expenses: Operating expenses also increased, mainly due to AI-related investments, including employee compensation (including stock-based compensation) and computing power depreciation. ⑥ Free Cash Flow: Although operating cash flow improved quarter-over-quarter, capital expenditures also increased, resulting in free cash flow of $146 million.
2. Policy Impact:
a. IRA Tax Credit Cancellation: The $7,500 electric vehicle subsidy under the Inflation Reduction Act (IRA) was canceled at the end of this quarter. Due to this sudden change, because of the lead time for component orders, our short-term U.S. market supply will be limited. This quarter, our vehicle supply in the U.S. is limited because we have entered the delivery cycle for ordering components to produce cars.
We have launched all planned incentives, and as sales progress, we will begin to reduce these incentives. If you are in the U.S. and plan to buy a car, please place your order now, as we may not be able to guarantee delivery for orders placed in late August and beyond.
At the end of this quarter, the $7,500 electric vehicle tax credit in the U.S. will be abolished, vehicle supply in the U.S. will be limited, and orders placed after late August may not be guaranteed for delivery, so please order as soon as possible.
(Dolphin Research Note: The cancellation of subsidies leads to increased car purchase costs, and Tesla needs to adjust production plans in advance to cope with supply constraints caused by the lead time for component orders. In the short term, there will indeed be a rush to buy, driving sales growth, but the supply chain may not keep up, causing some shortages or delivery delays.)
b. Tariff Cost Surge: Q2 tariff costs increased by about $300 million quarter-over-quarter (2/3 affecting the automotive business, 1/3 affecting the energy business), and the full impact will be seen in subsequent quarters, with short-term cost pressure.
c. Decline in Regulatory Credit Revenue: Carbon emission regulations will also be revised, with some penalty amounts reduced to zero, affecting our revenue from selling emission credits to other manufacturers.
3. Low-Cost Model 2.5: The low-cost model was originally planned for mass production in the first half of 2025, but due to prioritizing deliveries before the U.S. tax credit expiration and the complexity of ramping up new products, mass production will be delayed to the next quarter, progressing slower than expected.
4. Cash Flow: Q2 free cash flow was $146 million (operating cash flow increased but capital expenditures also increased); annual capital expenditures are expected to exceed $9 billion, as we continue to invest in Cybercab, Semi truck production lines, manufacturing processes, and AI engine expansion.
2.2 Q&A
Q: Can you introduce the current operation of autonomous taxis? What are the expectations for vehicle numbers, coverage, city expansion, and safety officer deployment speed?
A: Operation: The autonomous taxi operation in Austin is going well, with customers generally finding the experience smooth, safe, and satisfactory.
Coverage: The first round of expansion in Austin has been completed, and we plan to expand Austin's operational range to more than ten times the current size.
City Expansion: The San Francisco Bay Area is the next stop, with approvals underway, and services with safety officers will be launched first; testing is also being conducted in Florida, Nevada, and other locations.
(Dolphin Research Note: No mention of the time point for removing safety officers)
Q: What are the technical and regulatory obstacles to unsupervised FSD, and when is it expected to be available to individual users?
A: We are indeed approaching this goal. I believe that in some regions, unsupervised FSD services can be offered to individual users by the end of this year, but we will proceed very cautiously. This is not something we intend to rush; it must be completely safe before being widely launched. We remain highly vigilant about this.
It is worth mentioning that the hardware used in Austin's Robotaxi is the same as that in some user vehicles. We have even achieved automatic delivery of vehicles from the factory to customers' homes. Every Tesla vehicle produced in the U.S. and Europe can autonomously drive from the end of the production line to the loading dock, which is essentially a routine software issue.
We expect that by the end of this year, Tesla vehicles in the Austin area and the Bay Area will be able to achieve default autonomous delivery unless users choose to disable this feature.
Q: What specific tasks does Optimus currently undertake in the factory? What is the expected timeline for mass production and external sales? How does Tesla expect Optimus to contribute to revenue in the next two to three years?
Optimus 3 is what we currently consider the "right" design, although there will be optimizations in the future, no fundamental changes are needed, and it has all the necessary degrees of freedom.
Mass Production and External Sales Timeline:
Next Three Months: We will launch the Optimus 3 prototype and start small-scale production within the next three months;
2026: Mass production will begin early next year, with limited impact on revenue in the initial stages of mass production, as costs exceed income and many issues still need to be resolved.
However, considering that this is a brand-new product designed from scratch, the production ramp-up will be constrained by the weakest links in the supply chain and internal processes. Typically, mass production does not generate significant revenue in the initial stages, but rather involves troubleshooting and problem-solving.
Long Term: Monthly production capacity is expected to reach 100,000 units within five years.
Q: Can you update us on the development and production timeline of Tesla's more affordable models? How will these models balance cost reduction and profitability? What impact is expected on demand in the current economic environment?
A: Timeline: Production started in June, with a focus on enhancing premium capacity in the third quarter (prioritizing the North American market to maximize existing model production before the IRA policy expires); despite the imminent cancellation of subsidies, we will still launch more affordable new models in the fourth quarter.
Cost and Profitability Balance: The goal is to balance revenue and gross margin by creating affordable and market-favored products.
Q: Can you discuss the benefits of Tesla's investment in xAI?
A: This is not a topic suitable for discussion in an earnings call; if necessary, we will provide further explanation. We are a publicly traded company, and shareholders can propose and vote on any proposal, and we will respect shareholders' wishes and take appropriate action.
Q: Can you talk more about the work of Tesla's design studio?
A: We usually do not disclose the content within the studio, but it is certain that the design studio is working on many exciting projects. In the coming years, Tesla will transition from the "pre-autonomous driving era" to the "post-autonomous driving era," and I am preparing a new "master plan" to explain this transition. This process will bring some growing pains, but we believe that the future Tesla will have a positive and profound impact. If this vision is effectively executed, Tesla will become the most valuable company in the world.
Q: Will HW 3.0 users have any news about hardware upgrades or updates? Will they receive HW 4.0 or future HW 5.0 versions?
A: The current priority is to achieve driverless functionality on HW 4.0. Once the driverless functionality of HW 4.0 is completed, we will then study the arrangements for HW 3.0 models.
Q: Can you update us on the progress of Dojo? Is it possible for xAI to become a customer of Dojo?
A: Dojo Progress:Dojo 2 is expected to operate on a large scale next year, equivalent to about 100,000 H100 units; the AI 5.0 chip is expected to be mass-produced by the end of next year, with great potential; we are planning the integration of Dojo 3 and AI 6 inference chips, where a single chip can be used in cars, Optimus, or multiple combinations applied to data centers, making this architecture more intuitively reasonable.
Q: Will the cancellation of the solar project tax credit by BBV affect Megapack's sales channels?
A: Our Megapack sales channels are diverse and are less affected by solar project integration.
The market is increasingly focusing on the efficiency improvement of the grid through energy storage. Despite the adverse effects of this bill on solar energy, we believe solar projects will continue to advance because the demand for electricity is urgent, and the projects are already executable. Natural gas turbine delivery cycles are long and expensive, with no alternative solutions in the short term. Additionally, the demand for data centers and standalone energy storage projects is also growing rapidly. We expect deployments to increase significantly in the second half of the year.
Countermeasures (Localization of Manufacturing): At the same time, we continue to invest in local manufacturing in the U.S., with the first LFP cell factory set to start production within the year, and the third Gigafactory to be completed in Houston by 2026.
Q: Can you share the key performance indicators (KPIs) of the robotaxi business? I want to understand the progress of this business and whether there are any publicly shareable goals.
A: Key Indicators: Autonomous driving mileage in the Austin area exceeds 7,000 miles; since the service has just launched, the number of vehicles is still limited, but we are rapidly expanding the service area and vehicle numbers, covering Austin and other cities. Currently, the number of vehicles is small, and there have been no major safety incidents so far.
Progress: The service has just been launched, and we are expanding the service range and vehicle numbers in Austin and other areas; some restrictions (such as a speed limit of 40 mph) are set for convenience rather than safety considerations; the service has received positive feedback.
Q: In the long term, you mentioned reducing the cost per mile of robotaxis to $0.30-0.40. Now that the service is online, what are the main milestones for achieving this goal?
A: Cybercab is a model optimized for autonomous driving, with the potential to reduce costs to below $0.30 per mile, or even $0.25.
It is not focused on driving pleasure like the Model 3 or Model S, but rather on smoothness and comfort, reducing reliance on high-performance brakes, tires, and acceleration systems, with a more efficient and energy-saving design.
If Optimus robots are responsible for cleaning, maintenance, and charging, operational costs will be further reduced. Although the cost of existing models is higher, they are still competitive. We expect the Robotaxi business to have a substantial financial impact around the end of next year.
Q: Is Elon concerned that with only 13% ownership in the company, he lacks sufficient control to support Tesla's entry into the "physical AI" phase?
A: Yes. This is a concern of mine, as I have mentioned in the past, and I hope it can be addressed at the upcoming shareholder meeting.
But it is indeed a big issue, and I worry that I have too little control and could be ousted by aggressive shareholders after launching a legion of humanoid robots. As I have said before, I believe my control over Tesla should be sufficient to ensure the company moves in the right direction, but not so much that I cannot be removed even if I go crazy.
Q: I heard that the board of a major U.S. investment bank recently visited the Optimus production line. Can you confirm this? When will others have similar firsthand experience opportunities? Is it too early to hold an AI Day in the second half of this year (considering the intense talent war, previous events were important for recruitment)?
A: This is indeed a dilemma. Every time we hold an AI Day, competitors analyze our content frame by frame to copy it. Of course, AI Day is beneficial for talent recruitment. We are considering whether to add in-depth introductions on AI, Optimus, and chips at the shareholder meeting. Our AI chip design is actually severely underestimated, and there is still no AI chip on the market that can replace our self-developed chips for use in cars.
We believe this AI chip will bring about a huge transformation. In fact, its performance is so powerful that we have to impose some restrictions for markets outside the U.S. because it far exceeds export restriction standards. Unless export restrictions are relaxed, we do need to limit the performance of Al 5, which is a bit strange. Hopefully, export restrictions will continue to increase, otherwise, it would be ridiculous.
We will showcase multiple Optimus robots at the shareholder meeting, and the Optimus lab is cool, looking a bit like a scene from "Westworld." The robots are at different stages, with some still in preparation. It's hard to describe, a bit like a combination of Tatooine's junkyard and "Westworld." Very cool. Optimus is walking around the Palo Alto office 24/7.
Just like usual. I think Optimus is like the Tesla restaurant, gradually becoming popular. Yes, we will move from a world with few robots to one where robots are everywhere and people are used to it.
Q: When can personal vehicles join the Tesla Robotaxi network?
A: Our current focus is on ensuring the deployment of Robotaxi with full vehicle control first. This needs to be done step by step, especially as we are very cautious about safety. But I am confident that at some point next year, users will be able to choose to add or remove their private cars from the Tesla Robotaxi network. We will set certain standards, similar to how Uber or Lyft require vehicle inspections. This is more like an "Airbnb for cars." Details such as tire wear will affect safety, and we will establish reasonable review mechanisms to ensure all vehicles joining meet the standards.
Q: What is the cost structure of the Robotaxi business during expansion? Will the balance sheet be used for financing?
A: Once Robotaxi forms a stable cash flow, it can be supported through debt financing. Before reaching scale, we will use the company's balance sheet to transition. Once a certain revenue scale is reached, external funding can be obtained through simpler financing methods.
Q: What is the adoption trend of FSD (Full Self-Driving) subscriptions? What is the current revenue scale of FSD?
A: As I mentioned in the opening statement, since the launch of FSD V12 in North America, we have seen a significant increase in FSD adoption rates. Last year, we also lowered the subscription price to make it more affordable. Since then, FSD subscriptions have increased by about 45%, which is a very encouraging trend.
As Elon said, "seeing is believing," like that cat that can sing and dance. After the release of V13, FSD subscriptions in North America increased by another 25%.
Promotion Focus: We have just started explaining the value of FSD to users. Even if they don't believe in other content, the fact that FSD can improve vehicle safety by ten times is reason enough to use it. Additionally, even at a subscription price of $99 per month, it is equivalent to only $3.33 per day, allowing you to have a "personal driver." This is a disruptive change.
We discuss FSD a lot internally, but most people don't even know it exists, and even half of Tesla owners have never tried FSD. They don't know their cars have this capability. Most people are unaware of FSD's existence. We need to strengthen user education, such as reminding them when they visit service centers and sending them instructional videos. FSD is easy to use, but users find it hard to imagine that the car can actually drive itself. It's like a cat that looks ordinary but can sing and dance, but if you don't show it, they just think it's a cat. FSD is like that, it doesn't look different, but it's very smart. To address this issue, we will enhance FSD trial prompts, such as suggesting users try FSD during a certain drive.
Q: Will you use pricing strategies to drive sales growth, especially in the context of the impending cancellation of IRA subsidies?
A: We are currently in a transition phase, with many incentive policies being lost in the U.S. and other regions; at the same time, we are also in the early stages of autonomous driving development. Although the regulatory environment in the U.S. is the most open, we may experience a few "difficult quarters," possibly Q4, Q1, or Q2. Although I cannot be certain it will happen, it is possible. However, once autonomous driving is scaled up in the second half of next year, I believe Tesla's economic benefits will become very strong.
Q: Regarding the low-cost model you mentioned (production started in the first half of the year, with subsequent ramp-up), can you elaborate on its specific situation? Compared to the content mentioned at the last analyst day (such as reducing silicon carbide usage by two-thirds to three-quarters, not using rare earths in motors, and other cost reduction measures), what updates are there? Does this model use the unboxing production method that was said not to be a transitional solution?
A: We will not disclose the appearance of the model for now, but I can say that the demand for this car is very strong, but many people do not have enough balance in their accounts to afford it. This is not a matter of willingness, but a matter of payment ability. Therefore, the cheaper the vehicle, the better. In the future, when users can add their vehicles to the Robotaxi fleet and earn income, just like homeowners renting out their homes through Airbnb, the "affordability" of vehicles will increase significantly. I am confident that this will become a reality in the U.S. next year.
Q: The xAI project (such as Grok) is developing rapidly, and Tesla is also actively investing in AI. How do you coordinate the division of talent, resources, and funding between xAI and Tesla? Will there be internal competition?
A: The two companies focus on different directions. xAI is developing TB-level, or even multi-TB-level models, which belong to the realm of general artificial intelligence (AGI). Tesla, on the other hand, is developing AI models for the real world, with a scale only one percent of xAI's. In fact, one of the reasons for founding xAI was that some top AI engineers were unwilling to join Tesla because they wanted to work on open AGI research. They didn't want to do AI for cars. Therefore, I envisioned that if we established a new company, these people might be willing to join. Instead of letting them join OpenAI or Google, we decided to establish xAI ourselves.
Essentially, we offer two different options: one is superintelligence (in data centers), and the other is real-world AI (in vehicles or robots). Both are highly attractive technical challenges, but each engineer has their own direction they want to pursue.
<End Here>
Risk Disclosure and Statement of This Article:Dolphin Research Disclaimer and General Disclosure
