Q3 2023 Baidu Inc Earnings Call
Hello, and thank you for standing by for Baidu, <unk> third quarter 2023 earnings Conference call.
At this time all participants are in a listen only mode.
After managements prepared remarks.
Will be a question and answer session.
Today's conference is being recorded if.
If you have any objections you may disconnect at this time.
I would now like to turn the meeting over to your host for today's conference Joanne Lim <unk> director of Investor Relations.
Hello, everyone and welcome to Baidu third quarter 2023 earnings Conference call <unk> earnings release was distributed earlier today and you can find a copy of our website as well as on newswire services on the call today, we have Robin Li our co founder and CEO.
Our CFO. In addition, our EVP in charge of Baidu AI cloud with APG.
After our prepared remarks, we will hold a Q&A session.
Note that the discussion today will contain forward looking statements made under the safe Harbor provision of the U S Credit Securities Litigation Reform Act of 1995 forward looking statements are subject to risks and uncertainties that may cause actual results may differ materially from our current expectations well detail detailed discussions.
These risks and uncertainties. Please refer to our latest annual report and other documents filed with the ICC in Hong Kong stock exchange.
<unk> does not undertake any obligation to update any forward looking statement, except as required under applicable law. Our earnings press release and this call include discussions of certain unaudited non-GAAP financial measures. Our press release contains a reconciliation of unaudited non-GAAP measures to the unaudited most directly comparable there.
The comparable GAAP measures is available on our IR website at IR Baidu dot.
Dot Com as a reminder, this conference is being recorded in addition, a webcast of this conference call will be available on Baidu IR website, I will now turn the call over to our CEO Robin.
Oh Hello Baidu.
Baidu worked delivered solid revenue profit and cash flow in Q3.
Alright, navigating a challenging macro climate for both our online marketing business and E on cloud.
I'm proud that our team managed to strengthen operational efficiency and making stable margin.
A full scale, we mentioned all of our product portfolio.
Ernie and Ernie vault.
Today, I would like to share an update on the new opportunities.
Already bought have opened up for us after that I'll discuss some key highlights of each arpin.
Presently we are in.
Mr Ross.
Broad based sampling shifted driven by January to get paid.
And foundation models.
That is that's to revolutionize every industry.
On August.
31st we received the approval to deploy already bought on a large scale and open API to enterprise customers.
Since then we've witnessed a significant increase in inquiries handled I already bought and school learning API.
Moreover, we received valuable feedback from users and customers, enabling us to further refine our model performance.
At Baidu World in October.
We showcased our progress already bought and Nike product.
During that event, we introduced Ernie Volkmann theater or E before.
Our most advanced Foundation model.
We believe that E. Before he says TBD or level model displaying human level of performance in understanding content generation complex, we didn't reasoning and our memories for attention.
These capabilities are crucial for your balance sheet.
Application and solution.
We are pleased to launch E before earlier than our expectations.
This resulted from our unique end to end order here.
Infrastructure.
Which helps.
You should see in auto between.
Their input and feedback from our users and customers also played a big role.
In products as I said in the past, we continue to use our need to reinvent our entire portal ordeal and introduced a native experience.
On the customer front.
<unk> bought enables baidu search.
By generating direct answers to search query.
Can complementing traditional search.
In the quarter, we initiated tests on our new features that recommend newsfeed like information together that generated search results and enable multi wrong conversations to encourage further user expression.
Our initial tests have received promising feedback.
We believe this feature will help.
Deepening user engagement and prolonged time spend unleashing new monetization monetization opportunities.
In particular, they benefit verticals like health care education travel legal and auto.
Advertisers are willing to invest heavily in customer acquisition and engagement.
Only bought our new AI native prana.
As a person how multi pronged conversational AI assistant on both desktop and mobile.
Given the exceptional performance of our large language model, we are confident in monetizing our services.
Starting from November 1st E before with open to the public school earn you bought.
Our subscription fee of about eight U S dollars per month.
This marks the us as the first company in China to implement user charges distinguish us from other models in the market.
Our primary focus is to encourage seamless collaboration between users and AI co.
Co pilots.
Which we believe is a key twin England new era.
For example isn't until pilot Baidu, one cool has transformed into a one stop shop for whereas document creation.
We have already seen an increase in the.
The paying user account a trend that we expect it to continue in the coming quarters.
On the enterprise side.
Are you seeing products front.
We recently introduced a G behind generators benzene intelligence with Orange box GB.
<unk> simplifies data analysis, using natural language interaction facilitating faster decision, making for business operation.
The introduction of G. B I was prompted by this recognition that customers across various industries have the need for AI co pilot to help them in life data more efficiently.
During our last earnings call. We also discussed how we use already bought two eight baidu combing.
Our AI coding assistant and impossible or enterprise communication and collaboration platform.
This product forward.
Boosting productivity and efficiency gangs.
And each of them present upsell opportunities for our cloud customers.
Is that an increasing number of our cloud customers and China's traditional industries and public sector have used.
The cloud version and show interest in this product and features.
Additionally, this product's been features also allow us to acquire new customers across an even wider array of industries.
In terms of ecosystem, we empower enterprises to language learning suite API to create their own AI.
Applications and solutions.
That will drive the development of generative AI and L. L M.
As more and more native.
Applications built on top of earnings become successful, whether it'd be blocked by us or by our customers Army well likewise be successful.
Now over 10000 enterprises are actively using earnings through a P. I.
Lucky basis. This number has been growing quickly since we received the regulatory green light.
At the end of August.
Currently Ernie is handling tens of millions of queries every day right.
Right now a large and growing number of this quarry come from the Baidu family of products as we have been pioneers in beauty 18 products and have put a lot of effort into really a mid teen or offerings.
In the first half of November the number of daily external queries have increased by over 50% compared to the same period.
October.
We are actively assisting our customers in creating AI native applications. We believe there will be a continuous and significant rights in external players in the future.
We are also actively attracting developers to connect to their information and services to earn he bought scoop clunky.
With Pluggings Ernie can help people with more and more tasks allows them a wide range.
<unk> use cases.
As of today.
Pluggings have already been accessible to learn.
The initial batch of third party plugging into Ctrip Dot Com Citic press groups, China, just these big data Institute, New Oriental auto home and treat mind.
In summary during the quarter, we made significant progress using Jennie O I to revolutionize the product usage and transformed business the operations.
Our users and customers.
We believe that this is just as a beginning in the future we will realign resources to invest in these growth opportunity and shift away from lower priority efforts and improve efficiency for existing businesses, that's balancing investment and margins.
We are excited about the possibilities for baidu for our users customers partners and the entire <unk>.
<unk> system.
Now, let me recap the key highlights of each of our businesses.
Mobile ecosystem continued to.
It's a steady growth support both user metrics and financial performance in the quarter.
Baidu apps and they use increased by 5% year over year to 606 T screen Mellon in September.
Search queries and content distributed by Baidu App remained resilient in particular videos distributed by search and feed within Baidu App, both experienced double digit growth in third quarter.
Baidu course online marketing revenue increased by 5% year over year in the third quarter consistently generating strong profits and cash flow for the group.
This growth was driven by the continuous recovery in verticals, such as health care and travel among others.
England corner, we continue to use AI to help advertisers increased ROI and conversion on our platform.
Starting from September advertisers can engage with our new marketing platform.
This platform supports natural language input and multi wound that conversation, which help advertisers articulate their requirement more comprehensively.
Tabling us to formulate more effective campaign strategies for them.
Moreover, we made ongoing enhancements to our monetization system, focusing on improving targeting capabilities and the auction system.
For example to arena education.
I T professional education company achieved an increase of 23 point suite percent in conversion rate and 22, 7% boost in ROI after using this enhanced our platform and capabilities.
We are still in the early stages of using choosing eye to help advertisers achieve higher.
Conversions in our eyes on our platform.
Our efforts will ultimately lead to significant improvement in monetization capabilities and contribute to future revenue growth.
In the corner, we also used our AI capabilities to help merchants Rosie ourselves on Baidu.
One example is how we help ease these livestream shopping.
But it's early Bob we introduced a tool that allows them to easily create their own digital humana.
Generally in line with street lives scripts and more significantly lowering the barrier and cost for live streamers to sell merchandise on baidu.
Looking forward, we are optimistic that the growth of our online marketing revenue will continue to exceed China's GDP growth.
At the same time, we will continue to test it.
Our native marketing products that could potentially open up more opportunities than traditional general search ads.
This gives us confidence in Baidu longterm online marketing growth prospects.
Turning to AI cloud, we continue to generate positive operating profit on a non-GAAP basis in the quarter.
We remained focused on the health needs of Harbin.
Turning I M. L. L. N have brought us a lot of opportunities, which have strengthened our competitive advantages in cloud and increased our Tam.
A growing number of enterprises, using Ernie API two devices their own native applications and solutions.
We're also helping customers build their own models efficiency by leveraging our unique or New York.
Infrastructure and.
Our years of experience in beauty and using foundation models.
Oh, well cleaning is very complicated and requires a large number of gpus working simultaneously ending GPU failures can impact the entire process.
We have to go back a ways to identify and address G. P. You'll see there is quickly leading to a significant reduction in training costs.
No about 80 are about 98% of the training time on our platform is that it setting an industry benchmark.
We also have a set of different resources, including toolkit, a dataset for enterprise customers to easily fine tune their customized models.
Generally I has to help us grow our cloud customer base, a large number of cloud customers using oney API on new customers.
At the same time some of our existing cloud customers have increased their spending with us because of generally Dave AI.
And cloud revenues declined by 2% year over year in the third quarter, mainly due to the weak demand in smart transportation projects.
We believe AI cloud revenue should rebound to positive growth in the fourth quarter, driven by the increasing momentum being generated by AI related dependency.
All of those things Smart transportation revenue started to slow down in Q4 of last year, we will have an easier year over year.
Our comp base in Q4 this year.
Moving onto intelligent driving our target remains unchanged, which is to achieve breakeven on the regional unit economics for solar tax the operation in a couple of years before turning operationally profitable.
To this end, we are strategically concentrating our resources pivotal regions.
It remains our largest operational area and we believe it is also the largest region globally, providing autonomous ground hitting somebody says currently covering a population of about $2 7 million.
In the third quarter, the portion of fully driverless orders within the overall their portfolio.
Exceeded 40% that's up from 35% in Q2.
We are also particularly pleased to highlight that Apollo goes operation.
We continue to expand.
In late August.
Our political remains the first accompanying China to provide Brian Haney services to the general public S. One canceled international airport one of the busiest airports in central China.
The extended reach into the airport transfer involves longer travel distances.
Presenting an excellent opportunity for the future improvement of unit economics.
All of this development contributed to the improvement and we aim to reach reach break even in a couple of years.
In Q3, our Pogo provided 821000 right to the public.
Marquee 70 suite percent increase year over year.
And accumulate them all their volume.
Surpassed full Quinn 1 million by the end of Q sleep.
That's part of our executive we Shuffle program. We have recently named the top there I'm getting pumped as corporate VP of Baidu <unk> beat the intelligent driving groups.
In Poland has been with US 2012, and has been responsible for autonomous driving been inked 2018.
I take great pride in saying another been sliter deposit breathing baidu.
Jamie has taken a rotational position as CEO assistant and the chairman of the Technology Ethics Committee.
Now, let's proceed with schools financial performance Gravy thinker.
Think of it now.
Now let me walk you through the details of our third quarter financial results total revenue was RMB 64, 4 billion, increasing 6% year over year revenue from Baidu core was RMB 26, 6 billion, increasing 5% year over year Baidu.
Baidu course online marketing revenue was RMB 19, 7 billion, increasing 5% year over year Baidu cost not only marketing revenue was RMB six 9 billion up 6% year over year revenue for <unk>.
8 billion, increasing 7% year over year.
Cost of revenue was RMB 60, impulse throughput in which remained essentially unchanged compared with same period last year operating expenses were RMB 11, 5 billion, increasing 8% year over year, primarily due to an increase in China's spending promotional marketing expenses so about depreciation.
Expenses and and the silver.
Because of these pits, which support earning ball ruthless in course baidu costs operating expenses were RMB 10, 5 billion, increasing 10% year over year Baidu call SG&A expenses were RMB, four 8 billion, increasing 14% one four year over year SG&A accounting for 18% one pay off Baidu.
Called Rockville in the quarter compared to 17% one seven in the same period last year.
<unk> expenses were RMB, five 6 billion, increasing 7% year over year R&D accounting for 21% of Baidu core revenue in the quarter were instrumented unchanged. So long as then Purion last year.
Operating income was RMB six plants throughput in Baidu core operating income was RMB, five 5 billion and Baidu costs operating margin was 21%.
Operating income was RMB seven 6 billion noncash Baidu core operating income was RMB, six 7 billion and noncash Baidu core operating margin was 25%.
Total other income net was RMB, one 9 billion compared to the total other loss net of RMB four 8 billion for a century of last year, mainly due to the first a recognition of RMB 300 million gain versus RMB, surpassing 1 billion loss for the same period last year from favorable changes in the loan.
Investments and second a decrease in Panama for the long term investments by RMB, One 4 billion.
In context in context, but in context expenses was RMB one place for it but then increasing 14, 1% year over year, primarily due to an increase in profit before tax net income attributable to Baidu was RMB six 7 billion and diluted earnings tied yes, RMB 18 doubled.
Net income attributable to Baidu core was RMB six 4 billion and net margin for Baidu core was 24% non-GAAP net income attributable to Baidu was RMB seven 3 billion and non-GAAP diluted earnings played yes, well RMB 24 zero.
non-GAAP net income attributable to Baidu core was RMB 7 billion and knocking module for Baidu core was 26%.
Asaf sawtimber such is 'twenty, two 'twenty three cash cash equivalents restricted cash and short term investments well RMB 200, or two and <unk> 7 billion in cash cash equivalents restricted cash and shopping your investments excluding I'd see what's the RMB 197 point for Colette.
Free cash flow was RMB 6 billion and free cash flow, excluding I'd see it what's the RMB five times cheaper that baidu.
Baidu call approximately 5000 employees ourselves so dispersed.
As of September 30th 10 between the three.
With that operator, let's now open the call to questions.
Thank you.
We will now begin the question answer session.
If you wish to ask a question. Please press Star then one on your telephone and wait for your name to being out.
If you wish to cancel your request. Please press Star then two.
Your first question comes from Alicia Yap with Citi. Please go ahead.
Hello. Thank you good evening Robin Julissa management team. Thanks for taking my questions. My question is just on advertising.
So it it seems like if I do a revenue growth is tracking them slower than some off the internet peers. So besides macro can management elaborate them any other reasons that contributed.
Two the softer revenue growth and then looking into the fourth quarter I.
Have you seen any demand picking up why do you say e-commerce sector contribution and how well a I change the advertising outlook. Thank you.
Hi, Alicia this is robin.
<unk> suite apart from the macro weakness online marketing revenue from E. Commerce platforms was also relatively weak.
Revenue from e-commerce platforms.
One of our top revenue contributors.
It accounted for about 10% of our total online marketing revenue.
Like many other internet platform companies, we are building our own native e-commerce businesses.
Revenue growth from our native E. Commerce spend is tracking very strong as we continue to improve the shopping experience on baidu.
I would like to highlight the strides we have made in our Athena School Chin AI.
Basically a restructuring the overall ad platform.
Cooling creative construction, our AD targeting and the beating mechanism.
These efforts have started to pay off.
And does the incremental revenue from this kind of initiatives.
Got it.
To reach the level of hundreds of million RMB in the current quarter, which is Oh Q4 of this year.
And looking forward, we are optimistic that our the growth off our online marketing revenue, while continuing to exceed kind of GDP growth.
Thank you.
Yeah.
Okay.
The next question comes from Alex Yao with Jpmorgan. Please go ahead.
Okay.
Thank you management for taking my question I have a few questions on revenue.
Uplift Robin mentioned or despite a.
Moderate revenue decline in Q2 revenue.
Gross rates will return to positive territory in Q4.
And then from there should we expect cloud revenue to further accelerate into first half of 2024.
With regard to the Smart City project are there any more projects that are still at risk.
And then winning importantly, I'll see you guys start to monetize the I commit to you when we start contributes to the cloud revenue meaningfully lastly, any preliminary view on cloud revenue growth all of them.
For 2024, thank you.
Hi, Alex This is adobe. Thank you for your question.
So actually as we mentioned before so we have been focusing on improving the house for the business for sustainable development and as a result, you know we have achieved a non-GAAP operating profits in the past few quarters.
That's robin already mentioned right due to the weak demand for intelligent transportation and our cloud revenue experienced slight decline in the Q3 so.
So well, excluding the excluding the smart transportation.
If our alcohol business showed a pretty Saudi groups and.
And we believe in our cloud revenue will return to positive growth in the fourth quarter and the trend will continue you know down the road.
What is even more exciting side, we keep seeing new opportunities are bought up by a general G. P. I N launch Nangry modems.
Actually lost water water. We are already say that are you know more and more customers across various sectors Pinterest for.
Model training application development and solution enhancements.
Those are con revenue from the generally V I N L. L N no religious businesses do.
Very small, but it's growing very fast we have seen that more and more enterprises are proactively adopting these new technologies for productivity and efficiency gain.
Some of these customers, especially those from the unite education and the tech sectors. They have started to see our efficiency gains.
Working with Sears.
Yes, resolving some of them have gradually increased their spending on our cloud services.
Looking to Q4, we aim to leverage our leadership in the general G. P. I N G. Large language models to continue to attract new customers and encourage the existing customers to increase their spending on that baidu a clock.
And we believe this should not only lead to a long term revenue growth, but also continuous margin improvement.
Thank you Alex.
Yeah.
The next question comes from Miranda Zhang with Bank of America. Please go ahead.
Thank you good evening, Thanks management for taking my question and congratulation on the result on my question is about early on so can management share with us very good feedback or any for us just a little I was last month.
Any color on the contract right Oh adopting Ernie for.
Oh, so for consumers and how is the feedback after any books starting to charge a M E series a subscription fee.
And lastly, among the various opportunities you mentioned, which one do you think can become the biggest revenue driver. Thank you.
Uh huh.
Let me answer your questions.
The release of E beforehand.
October.
Receiving positive feedback from both users and our customers.
Many of the prices have reached out to a test E before and have been impressed with the with its capabilities.
E. Before has gained a reputation for its advanced understanding and complex reasoning abilities.
Comparing to <unk> five and other out in the market. We have noted that our E before generally more structured and Sierra responses in itself in coding.
From November 1st and we started to charging in Nebraska and end users for using E before and we've seen a growing number of customers and users are willing to pay for its use.
We're proud to be the first company to introduce a G. P D four level model in China.
Before further widen our lead over other hubs in the market.
And we are the first out.
Two charts and user fees.
That's that's not a part from our other peers.
And regarding your question about monetization opportunity we.
We see significant opportunities in native applications, either a device or by baidu or by our customers.
We'll leverage our AI capabilities.
If you look at our own products, we see significant opportunities in the new search and revamped and platform.
The new search complements traditional search it can address complex questions that were previously answerable.
It also enables users to conduct a more personalized and Inc research, whereas topics and projects.
We will soon.
Enable users to have multi around the conversation with us.
F. Our search well be able to talk with new sourcing natural language any multi wrong it well.
Create more potential on the commercial side too.
Uh huh.
Experimenting with a chatbot type AD product for S N E and the brand.
We believe this will not only help drive effective conversion.
It also allow us to eventually transform from a CPC model to a Cps model.
And at the same time our capabilities.
Or are there as well.
Help our advertisers to better operate their bins baidu.
Our ongoing efforts to revamp the AD platform have already shown positive results and we will continue our leveraging generating them AI and.
Too.
Assist advertiser team.
Sheathing.
Durable are wide world with somebody.
In terms of empowering our customers with Jenny I as Don just mentioned our customer needs are different now.
Customers that do prefer to train their own model, but the GPU export restrictions well put the brakes on that.
It will eventually become clear that our training L. L M from scratch it is.
Very difficult.
Especially when trying to achieve emergent abilities.
So we will turn to advanced L. L M available on the market like Ernie for developing applications.
As customers become more advanced and using.
N two.
Create applications and more AI native applications powered by Ernie become more widely used.
We should be able to see continuous revenue growth school model in France.
Over the long term influencing should become a major source of revenue forward.
Meanwhile, we will also help customers to fine tune, our existing model offering to suit the art customized spending.
In each scenario.
Of course, our model, a better faster and more cost effective.
So in summary.
Jimmy I, and Oh, Wow, and well bring us massive opportunities we have already made good progress in commercialization so far and this is Justin.
Bikini.
A promising future.
Thank you.
The next question comes from Gary Yu with Morgan Stanley. Please go ahead.
Hi, Thank you management put opportunity to ask question can management share the latest advertiser feedback on.
On the AI powered at system upgrades and how do you think about the level of revenue boost to the core ads in 2020 full as it gets rolled out to more even or advertisers. Thank you.
Okay.
Yeah, we are very happy with the rapid transformation of our AD system and thrilled by the positive feedback from our advertisers.
Overall, I think advertisers appreciate our efforts to help them improve their whole lives on our platform.
Oh, so as the found off our new features which help them to be.
More productive.
As I mentioned in the prepared remarks.
We've put a lot of effort into using chain AI and ml to reinvent our apps system over the past few quarters.
Now we have an integrated advertiser facing marketing platform.
Advertisers can use it to generate a.
Creative advertising materials.
These materials have proved to.
Deliver a higher conversion rate than materials created by humans.
And our platform also allows advertisers to engage in natural language.
By interacting with the advertiser in multiple rounds off conversation our upgraded platform is able to.
A better understand their intention.
This allows us to create a campaign strategies that are deliver higher ROI.
Moreover, it significantly reduces the time that managers need to spend creating campaigns.
Even experienced AD manager have to spent ours goodbye.
Being an advertising strategy.
And now with our AI co pilot of the process only take like a few minutes.
As of today, we have a few thousands of advertisers have already migrated to our new platform.
While this number is relatively small compared to our.
Half million Advertiser base. It is certainly growing very fast.
In addition, we continue to use AI to improve our bidding system.
And AD targeting capability.
Oh, it's actually initiatives that happened at the backend of the system. So advertisers may not have directly perceive them, but the depot observed improvements in the AD conversion and all right.
So.
Al reach advertisers using.
This capabilities.
Probably achieved a high single digit increase income conversion.
In Q swing.
Okay.
All of our efforts should eventually.
To advertisers to allocate more of their advertising budget on Baidu.
Yeah.
As I previously mentioned online marketing revenue related to our upgraded our AD platform has been growing rapidly and.
Already became a meaningful.
And.
This is just the beginning as I said, we are experimenting with AI chatbot.
And chat Bots could act as a replacement for landing page in the future.
This is particularly useful in verticals, where users typically research and engage.
England's long decision process before purchasing.
Imagine searching for training program and being directed to a bot instead of a webpage.
With our AI chat box users can quickly learn about a brand specific information.
Product details and other stuff.
Although.
I think the multi around the conversation.
It's also a very.
Engaging.
And for advertisers to help them to stay.
Stay connected with potential customers and.
At key decision, making point, leading to better conversion and even direct sales.
We believe AI chat bot could work well with no kidney eye powered new search and bring it with me.
More opportunities.
You mentioned earlier that the incremental revenue from this.
This kind of initiatives.
Expected to reach the level of hundreds of millions of dollars RMB.
This is kind of a corner two four and ER visits certainly growing very fast.
Twinned.
Should continue and.
Further strengthen.
The year of 2024.
Thank you.
The next question comes from Wei Zhang with UBS. Please go ahead.
Good evening management. Thank you for taking my question I want to follow up on your cloud business. So could management share more color on how to think about the industry competitive landscape in China called market, especially among the internet cloud vendors and as well as when competing with a telco.
And also with the development of <unk>.
Generally with AI and a large tax model, how do we assess our competitive advantage against peers do we expect the competition to intensify next year as other companies try to make efforts and catch up with us. Thank you.
Great question. This adult a few actually alrighty notice right. The traditional clubs business is slowing down well generally V. I N launch Nangry motto is stepping in and are reshaping the competitive landscape of the.
Cloud business or industry.
In the past right the focus in the cloud market, what's actually on the us.
S, which is like a commodity and people are competing on that pricing.
But now with the rice, if a generally AI and the lunchtime intermodal as things are changing.
There's a growing interest among cloud customers.
Turning to Baidu to utilize these are sophisticated technologies to increase productivity and efficiency.
It came tourists you know not only because we have the most advanced AI technology.
But also because we have experience and track record and are using AI to help enterprises to solve problems.
There were some of them are still in the product experimental stages.
The high firm belief in the new technology to rebuild their products and services.
Because they have seen.
You know successful stories and overseas.
That is why we are seeing the new technology is increasing our time and.
Expanding our competitive edge.
So I said Robin mentioned it before is China's first JBT forced out model.
He also shared that the positive initial feedbacks, we have garners for E before.
Currently our team or our engage the nerd dialogues with our claims assistance I mean understanding the technology.
And utilizing our ernie to redevelop their existing products.
And create new ones.
So I can see that you know Ernie has already helped earth.
Attracting new customers and additional spending from our existing customers.
So here I would like to briefly bring up two points for our advantage compared it to the players in the market.
The first one is our unique four layer.
The infrastructure.
Gabe as yours are.
Flexibility to make adjustments or innovations at every layer.
To be able to be compatible with other leaders to keep driving efficiency in both model training and inference.
And the second one more specifically is our capability to develop GPU networks or clusters for a lot longer and more freely.
So this is robin just siding with 98% of the training time on our E infrastructure is valid.
As a result of our customers, including several leading unite and tech companies for increasing their investments in our service.
Furthermore, we will continue to leverage our unique advantage architecture.
Drive efficiency gain.
It will have first to a greatly reduced cost in our motor premiums and it brings our clubs.
In Europe, the flexibility to offer more compelling prices to our customers and further discounts in our competitive edge in the market.
Okay.
You know.
Regarding that petition AR from a telecom operators.
Yeah.
I would like to highlight our focus is on different market segments seems weak.
Differentiate ourselves with our yeah, a couple of days.
And particularly you know Ernie because we had mentioned.
Actually it's worth noting that you know we can coverage and we are actually come breaking cooperating on the many objects are projects.
As for other Internet companies in China.
So our strong El Cabo is and Ernie.
Oh, well wix and I spend market.
Which will set us apart from our peers.
To sum up you know, we believe that our strong capabilities and particularly in a generally high end luxury models.
Allow us to eventually become the go to.
A market leader and gain share in the cloud market.
The next question comes from Lincoln Kang with Goldman Sachs. Please go ahead.
Thank you management for taking my question most of my questions Oh, so far to Ernie.
So given the successful upgrade after Ernie for zero, what would be the future strategy for model duration to certify our tech leadership. So do we foresee any competition and the foundation model and industry.
A bit too stiff when I saw it intensified in the future. Thank you.
Huntington.
The chips, we have in hand already allow us to launch E. Before we head off that competition.
To take her lead.
And to the next level.
We will take an application driven approach.
We will let the AI and native apps tell us what to improve.
Ernie Bot capabilities.
Given that there are only a very limited number of native apps are on the market right now the majority of our new API call from internal apps are internal apps like search as a wet and cool et cetera.
The rebuilding and restructuring of our existing products drive innovation in the right direction.
What is equally important is that we are helping enterprises use ernie to build their offerings and we have seen that over 10000 enterprises are using Arnie <unk>.
API cost on a monthly basis.
Now, which.
Propels our earnings improvement too.
We are also continuing to improve the efficiency of our model for example, compared to early box version. They learn your bond version during March.
Insurance cost off the current version has been reduced by 98%.
Basically rebounding in 50 times increase in Q P F.
For the same amount and that's all the hardware all computing power.
We are also April.
To do this.
Yeah.
Yeah.
Using our unique <unk> architecture.
And.
Hmm.
Leveraging our ability to do end to end optimization.
Continued increments cost reduction has further strengthened our model is a competitive advantage and it gave us the flexibility to offer more and more compelling pricing.
Now from a long term perspective.
Taking into account factors such as the viscosity of high performance chips.
High demand for data.
The talent and the huge upfront investments.
Industry will soon translate into a consolidation stage.
We believe there will only be.
Select a few foundation of all of those things in the market and Baidu, well certainly one of them.
At this stage of.
Industry environment.
More and more enterprise sits well begin to leverage advanced foundation models like Barney too.
To create.
Products rather than spending.
Spending resources on building their own large language models.
So we expect that the number of native apps based on Ernie.
Well reach millions.
In the future.
Thank you.
The next question comes from James Breen with Mizuho. Please go ahead.
Great. Thanks for taking my questions can you guys, maybe quantify the investments related to AI and how would that affect various cost items in your P&L and should we expect these investments to accelerate over the next few quarters.
Was thinking, especially after the launch of Ernie 4.0, and potentially higher insurance costs as more people using it.
And then if we extrapolate that over a longer term how should we think about baidu core O P. M. Over the next few years given all the moving parts, including you know revenue ship investment AI and Oh, So you'll continue improvement in cloud profitability. Thanks.
Hi, James limit their questions to this is true.
Currently the primary investments such under which the large engine models are centered around the computing power.
Wages recall, there is a powerful capex.
I think the past few quarters, we have pulled a lot of chips resource it's too.
<unk>, our new earning models.
And in the future us more niche applications, which is powered by our earnings on a more.
Why they did use a wish him well.
It was also more resources towards it.
All of US have just no doubt all office and powerful AI related to how we our investments all P&L is quite manageable.
Because the all the highway depreciations Osprey, all over last two years.
For example, at least all expenses being shoot a quoting power when you all seem to training Ernie I recorded.
Sure <unk>.
And the model inferencing causes ratios highly related to the usage of the models.
Do you guys internally or externally and you should be supporting all of them will be funded by a left.
Fueled tributes developments.
Moreover, this where we're happy to see the average investment.
General.
And lots of empty models are beginning to bear fruit as Robin has mentioned just earlier.
We're receiving the approval from the regulatory and.
I think from the revenue generated from earnings power bi to see all <unk> businesses has been growing.
Quite fast.
While we're using that.
And a lot of times your models to run our businesses, we are still keeping a close eye on making sure our baidu core earnings solid.
In Q3, we can see that mobile ecosystem continue hype module.
Joining a very strong generation of cash flow.
Copies as continue is heresy gross and achieved profitability one more time.
And look my hat.
Wastes that the traditional clubbing is to remain quite profitable and a new opportunity arising from each other.
And last name to models.
So expect to have for a favorable margin of 110.
For intelligent driving businesses, our long term growth opportunities, we'll continue to invest waste masher pace or <unk>. However, you will.
Concentrate our resources.
Advocating them thrown now noncore businesses true religion, you ship.
Sure.
The related businesses all of this well.
I think vishal, so our long term goals.
James.
The next question comes from Thomas Chong with Jefferies. Please go ahead.
Hi, Good evening, Thanks management for taking my questions and congratulations on a solid set of results.
My question is on the chip side can management comment about the impact on the AI you bought them and after the food that'd be stretching halt the chip export OSM to glass.
How does that affect our AI product offerings and do some speedways are if any thank you.
Yeah.
Yeah. So the restrictions on the cheap export to China actually had limited impact on Baidu.
In the near term.
We have successfully launched E before in mid October.
Our most advanced foundation model in China.
It is milestone for us and as.
As I just said earlier we have.
Substantial reserve of AI chips.
Which can help us not keeping.
Keep improving earn you bought or the next year or two.
Also influence.
Requires less powerful chips, and we believe our tip reserve.
As far as other alternatives will be.
Sufficient to.
Support.
Lots of native apps.
Floor.
The end users.
And in the long run having.
Having difficulties in acquiring.
The most advanced chips.
Inevitably impacts the pace of AI divestment in China.
So we are proactively seeking alternatives.
Well this options I'm not.
Not as advanced.
As the best chips.
The U S are unique for Lee, our AI architecture, and our strengths in AI algorithm.
Well continue.
To help us improve efficiency and.
Ah mitigate some off system challenges.
For example, we have made some innovation theme pedal pedal or deep learning framework and Ernie are foundation model to allow them to.
<unk> be better compatible.
Different pipe stockpile.
AI chips.
Both the motto training and increment that task.
But given that all of the other Chinese companies face the same challenge.
We believe we are actually faster.
Best positioned.
To serve this market.
As you can probably knowing the past.
Some of our peers they they try to write up the chimney I wave by investing those start ups.
Two train foundation models.
And.
They they basically sell the computing power to the startup.
We didn't do that.
We try to optimize.
Have you seen from the infrastructure layer or two.
Framework layer and to model. They are then to app. So we invest it.
In this kind of end to end optimization.
Approach therefore, we can hope.
For the same my mountaintop computing power, we can do training.
More efficiently more cost effectively and we can do inferencing.
Faster and cheaper.
Cheaper and buy.
As time passed by I think more and more company well realize that they don't need to train their foundation model if they they just need to do buyouts with native apps.
Based on Baidu with.
Our nation model, which is the best on the market. So I'm really happy that we.
We basically invested on this.
Kind of end to end optimization front.
Many many years and it's time for us to show that Oh, that's the investment base.
What's that.
Thank you.
Ladies and gentlemen that does conclude our conference for today.
Thank you for participating you may all disconnect.
Okay.
Okay.
Okay.
[music].
Okay.
[music].
Okay.
Okay.
Yeah.
[music].
Yeah.
Okay.
Okay.
[music].
Okay.
[music].
Okay.
[music].
Okay.
Yes.
[music].