Baidu (BIDU) 16 May 232023 Q1 Earnings call transcript

Hello and thank you for standing by for Baidu's First Quarter 2023 Earnings Conference Call. At this time, all participants are in listen-only mode. After managements prepared remarks, there will be a question-and-answer session. Today's conference is being recorded.

If you have any objections, you may disconnect at this time.

to conference, meeting of Juan like host to Investor now today's for Lin, would turn over the your Relations. I Baidu's Director ahead. go Please

Hello, on can find conference copy XXXX on was everyone. as first a earlier Baidu's well website, as distributed you earnings release services. Welcome earnings and today, our Baidu's newswire to call. quarter

and our the Zhenyu Cloud Luo, our Rong call and Co-founder CEO; AI in-charge SVP today, On our Baidu Dou Li, we EVP Shen, of our Group, Driving. of have ACG; Robin Li, Intelligent in-charge CFO; Baidu

remarks, session. U.S. the the prepared will safe the our contain we note today XXXX. of provisions Private After will Securities hold Q&A discussion under harbor Please Litigation a forward-looking Reform made statements of that Act

risks results and cause current to may our from subject to materially that actual expectations. uncertainties are statements differ Forward-looking

discussions our to as to with Stock applicable For any not please risks documents latest law. required SEC undertake annual these Exchange. and detailed uncertainties, filed statements, and except Hong any Kong forward-looking other does our and refer under of Baidu update report obligation

call press discussions earnings this financial include and release non-GAAP of certain measures. Our unaudited

on comparable Our measures measures to directly unaudited press and reconciliation non-GAAP contains unaudited at most is release the the ir.baidu.com. website a of our available IR GAAP

As conference a [is reminder, this being [ph]. recorded]

a webcast will be available conference addition, of In call website. this on Baidu's IR

CEO, to Robin. now will our I the call over turn

across the growth driving supported off is revenue XXXX our improvement and operating and to a performing good board everyone. an margin. well our growth. Hello, Baidu year-on-year with In start strong reaccelerated Core's QX, in business

local our in such quick which improvement. After many for impact positive there was and Our economic especially as services. bin the overall was seen beneficial online Year, marketing a has healthcare, services, sectors advertisers China's business Chinese offline travel, recovery, New a from of

previous to led the good working online quarter been compared revenue enhance and metrics for marketing to apps first to and year. user we've this Baidu the experience hard user for has increased Moreover,

AI quarter. on a last non-GAAP operating projects out cloud Our this have on focused sustainable quality Cloud building for phasing level QX our in low we business. by and Starting growth profitability achieved year, business

At project profitability them This building led in operating AI another. used on standardized approach non-GAAP have one and to replicating over seen from quarter. cases quarters key AI time, same the and to an the the has solutions margins we we Cloud past few level this been have increase for

impact all Today, transformative of keenly on eyes are AI. focused the generative

Intelligent operational I provide topic on AI in today's Cloud, will mobile ecosystem. then highlights first update an Driving, this through you and For and our call, walk

tremendous which bots capture years the model our our became achievements operated This engine. GPT and XX company. introduction it opportunity highlights in dialogue. our extensive AI the past opportunity, largest first can to We early able in operator be Generative launch and demonstrated represents position a search owned significant China's experience compared knowledge over for the mid-March, and AI also this of smartphones. Internet the graph, To technology public as our by will capabilities the leveraged and deploy were a us. quickly search, investments we It and and in of to

by critical With range of of the knowledge encompasses architecture, Ernie, robust understanding, capabilities which a we its PaddlePaddle our and amount cloud and for search established AI vast large in have models, has applications. natural has facts, a infrastructure, four-layered Baidu our AI data, framework, which engine, language models. deep-learning been AI language gathered powered training

optimize that in advantage helps and costs model us distinctive have higher for process reduce training and We this inference.

China. leading One deep-learning framework advantages owning of which PaddlePaddle, the significant is our in is

Bot. that these aligns a systems. in synergy strong the And developed PaddlePaddle a between this have way with We ERNIE creates

can which language AI architecture. give models Training computing stack use of example. be complex full demand resources. proprietary an To And you large we this running and tackle our a overwhelming. challenge, me and Let data, require significant lot

looking throughput. make better layers to achieve continuously new helps quicker We are our for techniques four work and which performance seamlessly, us

improved and PaddlePaddle, model, the the have language between Infrastructure. compatibility we Cloud large Specifically,

launching a testing. in Bot we efficiency As normalized almost significant per our gains. is – of queries have indicating ERNIE all our result, XXx, Bot ERNIE integrating Baidu businesses rapidly since March, second increased normalized for into

around sectors market We of in gain to The should products generative goal sustainable consumer AI is strategy and used to enable across our enterprise wider services, This reviewed business.] both working and thereby ERNIE Bot building us adoption and achieve the [our all an and on facing to ecosystem share leverage cases. to [ph] promote a range growth. of enhance long-term its are capabilities also revolutionizing create

people tailored to more real life seamless use ERNIE feedback integrate able ERNIE Bot As large products scale and more the gather be ways, human iterate to Bot, we will need. model and in rapidly market

of On solutions generative This the cloud productivity offerings, efficiency. or use providers the drive industry. total applications to AI market current the service [ph] and cloud the is capabilities enterprise generative unfulfilled and to expanding add facing and new [fin] develop demand because enhance addressable address can their AI side, to is market

training since a more from on our started various introduced have Bot, industries result, more their As we observed ERNIE Cloud. and we enterprises model

AI us. unique which market. by market to over which more share for increased in March, differentiate that Our AI gives full achievement from XXX% powerful Cloud a our In We competitors, continue is the capabilities lead will believe us sales gain foundation our significant edge us time. models than an and our stack year-over-year,

help the for employees this, we a industry, to domestic gain their I'll provide cloud. In insurance partnered prominent achieve company our ERNIE experiences PaddlePaddle and utilize capabilities, on digital Bot them of their clients. and AI some how with our to enhance examples services To we have Ernie private will including helping is financial customers efficiency.

efficiency improve models various we system create customized to refining power data. and continuously the Bot Additionally, time. reports sector. them unique advisors, assist that financial training believe financial will and These research ERNIE innovation standard models in will China's technologies analyze to of their and applications, in establish we tools amount on over with industry We information. And new will project public services including and a this the potential has for large helping

In the the applications to with their are software of ERNIE end developers automating certain of of goal users. with some productivity Bot and software the into the test collaborating their we industry, capabilities integration workloads expanding

assist Additionally, who partnering programmers of companies ERNIE writing some we Bot use are ERP exploring with their are code. in the to

model opened a ERNIE the In service, Baidu AI company model industry, education them own we as helping to leading online their via Bot fine on Cloud. tune a education

sophisticated their to students. We've on them that to are serve we're features surge the an in AI industry more of business better enabling One after working together tutor their experiencing operations using demand a Bot improve observed efficiently. pandemic is interested ERNIE the enhanced is in

companies travel good example are of Online this a trend.

ERNIE their As so and use suggesting customer beginning managing in switch service to The customers are customers some to far positive of to directing trips, assist cloud. and expenditures has their testing are teams customer to the of Bot the feedback been also biggest players queries, our information. relevant IT

conversational a by consumer beta search for users. of product businesses, search, enhance is version natural ERNIE our Bot powered This Bot. an currently testing more provides to On utilizing ERNIE experience. seamless interactions Baidu version we upgraded We're for which the facing are

assist summarizing, Bot Wenku, testing online generating our We searching, to to enable are users ERNIE content and alongside also is document integration language. sharing Bot natural with using work the Baidu of platform and ERNIE

for as Additionally, Bot. a we we ERNIE testing are Baidu such Post, are other and ERNIE into facing Furthermore, consumer application keyboard, standalone incorporating Bot products Baidu Baidu Map, Xiaodu.

look we forward far, the to So user have the encouraged more bringing stage to these experiences at been the users months. feedback in testing by and coming positive current

In seek to and around our mind Therefore, gain will to the an share, user traffic should will monetization working continue fuel time. new and we become for We're a traffic people partners era, to ERNIE build drive with Bot AI to ecosystem innovation which drive content. growth. chatbots also over information gateway and

or regulatory undergo the AI. AI and and AI must comply deployment. Baidu a large in promote before review as new the technology product continues ethics, requirements, to service On use front, regulatory responsible advocate government an with scale of any To leader approval for

have customers scale out a Bot green We we the get approval and at and will for our once large ERNIE users applied of review to and light. roll it

early operating will help and deep believe April, [AI for extensive more for the of cyber has China with regulation and XX which new In Baidu regulatory navigate on space explicitly environment, administration self-designed in experience [ph] development has years endorses the we of China than frameworks. which regulatory draft us released landscape. a algorithms] learning AI, search generative the Chinese comments been culture and

prospects that we invest well-positioned and regulators that in about envision. stage entry ahead believe in Overall, Ernie to in with we will the it thrilled continue the We the bar raise to to and active we early that potential AI tremendous are unlock will are generative for that. the we Bot, lie engagement

Now, look that profitability which operational let's last improvement margin Cloud quarter The during X% an Cloud We're the the [ph] increased is from the year-on-year, progress. quarter AI a have [unrolled] the quarter. AI X% profit fourth In to from highlights. growth QX, on points improvement level year. first of on operating in year-over-year pleased report two revenues the achieved non-GAAP

and efficiently to ability different growth. to healthy replicate AI solutions operations efforts and is sustainable and our standardize our is sectors. across optimize and One applications Two customers our drive revenue and them

margin industry, low a and we for to the of usage solutions key pressure structure. Since for developed Water the businesses for company replicated Changzhou used In water solutions to leading time, business same AI and replication. AI year, to example, year, phased for At projects portfolio out standardize from cases company. AI healthier last have last Cloud water controlling product and for water cost predicting QX we Quanzhou we continue we water

demonstrated and creating the expansion standardized majority our we standardized Gansu Province. Water traditional for scalability Solutions QX, industries. in dissolutions of to of deployed Company This AI approach County in the the Lintao earlier example our in And margin

AI our continue services. to is we the to invest As and I host of creating significantly. earlier, grow explore the will generative market Cloud mentioned for ways and Therefore, expanding opportunities Cloud actively industry cloud a

intelligent into driving. Moving

fully make reaching to the public rides continue period over rides government China. solid road] increased nearly ride to services in to track In growth the our same the offer of have driverless year the permits [ph] QX, hailing operation, [open number threefold, more in last provided We we and received on from to of and to Thanks strong progress. safe compared remarkable XXX,XXX. record

region was XX, Yizhuang Apollo Go allowed to driverless ride-hailing services fully On provide in of Beijing,. March

As of are services Beijing, cities, and Wuhan, providing Chongqing. the today, we three in

and passenger adoption fleet charge. fully it's driverless where We Meaning we can and in regions driverless prioritizing operation. can are provide commercial expansion we

yielding efforts Our are good results.

than one-third last provide received In services fully public vehicles more to permit have provided driverless the on Wuhan, lease more such ride of than where now first travel the vehicles. in we we growth were paid and XXX Go Apollo August,

the are to We by increase confident provided driverless cost continue business profitable and that number a XX optimization leading vehicles model. of creating to will rides

believe and better provide the firmly pace services for driving dedicate grow will demand and the public. capabilities our operational technology to We rapid continue services and further will we a autonomous at to to safe reliable

Moving positive ecosystem. a report reopening. year-over-year environment mobile increase One economic marketing QX, growth China's in as to that in factor to to pleased commercial returned China's I'm growth of and to overall improvement contributing an result revenues was online the in operating in this margin. recovery resulting

offline healthcare, increased. services, many and earlier, mentioned I have local substantially verticals, travel, services business As related

our at spending are on We platform.

recovery. indicating levels, of these rebounded pre-pandemic of above have Some [solid finds] [ph] verticals already to

Our own revenue app, videos within also initiatives such the at role a plays and product Baidu significant short as improving listing the aimed product increase. information the experience in incorporating

our facilitated grow, our GMV from on in the The traction to Additionally, continued grew revenues driven spending year-over-year e-commerce by quarterly mainly e-commerce platform. by QX, demonstrating merchant, XX% and sector. by Baidu retail in Search, continuous direct

and increased that Search The source queries mobile Baidu their users the steady through users and information. to by X% March search primary in on high MAU Baidu the continued quarter. year-over-year grow Our growth. app placed value and maintaining of traffic feed app distributed in as reflects Baidu This

user enhancing experience the we customer satisfaction. to continued and During on quarter, focus

side, We cases user results. page] incorporating the to wonderful Baidu diverse immersive [Explore app needs. On which feature used experience female, enabling is catering including [ph], app, on to infinite and browsing and attractive Baidu search video to more young an has our on it into expanded the flow more a particularly their

to improve increase continuously result AI These and through one-shop expand we helped efforts experience. to provide search better engagement Additionally, base. user have our search user our

customized specify options been By their the advertiser our generates advertisers requirements, create year, from. launching to side, capabilities our late platform for On since AIGC last more working we've advertisers with choose allowing effective them to ads. to

agency AIGC instance, to travel For ads. our create local a turn to capabilities

quarter-over-quarter In advertiser. ads higher by adopting outperformed sales business our created XX% their in technology those our AIGC and of average ad platform, advertisers about the generated increased leads. of AI effectiveness the driving spending on Delivering an Our QX, year-over-year demonstrating results.

we year, using their began an inquiries respond harnessing on our lead. can in advertisers which to turn our AIGC of and customer power customer the sales AI, drives this professional to our capabilities inquiries, Earlier answers increase to platform. help advertisers timely in By provide

For instance, using a our a XXX% advertiser in leads March in over witnessed increase of monthly after AIGC BXB technology. sales

have the on increasing them These results. improved business platform. helping their advertisers better ROI to service advertisers our in customers of for We and number are expanding efforts serve drive resulted improved

AI their target generated technology. by Looking ahead, to assisting our committed edge advertisers our converting we and cutting audience leveraging reaching remain in

me to through to With turn over Rong the financial call results. go let that, the

Robin. you, Thank

details let RMBX.X Total was increasing was quarter billion, from year-over-year. X% marketing was Revenue increasing up RMBXX first revenue through non-online results. billion, Baidu XX% Baidu year-over-year. our marketing financial revenue Now year-over-year. RMBXX.X me XX% you was increasing Baidu RMBXX.X the Core’s Core's billion, of online revenue year-over-year, X% walk billion, Core

in increasing an year-over-year was Operating In were year-over-year, X% market cost promotional was revenue decreasing decreasing [ph], iQIYI increasing revenue RMBX.X Cloud billion, Revenue of billion, X% year-over-year. to expenses. for of AI by Baidu billion] due X% and to billion, year-over-year. spending year-over-year. X% billion. RMBXX increased Cost RMBX.X [RMBX.X was RMBXX.X primarily increase QX, China expenses XX%

expenses SG&A year. increasing X% last year-over-year. period operating the Baidu Core billion, Baidu quarter, were to year-over-year. billion, XX%, revenue for expenses RMBX.X XX% RMBX.X increasing accounting were SG&A Core’s XX% of the compared Core’s same Baidu in

RMBX.X were last RMBX in Baidu was period Core for XX% from income billion. and Core RMBX.X year. billion, same Core’s non-GAAP Baidu decreasing the RMBX.X accounting quarter operating X% Baidu and was was revenues XX%. RMBX Baidu income year-over-year, income XX% margin Baidu billion. expenses in R&D operating R&D Core's decreased operating Core's Core the was Baidu operating income non-GAAP and Operating of and was billion XX%, was margin Non-GAAP billion, operating

billion first Total was of RMBX.X including of year, gain significant of RMBX.X fair under other billion, from mainly value a portion the value investments. last loss RMBX a net we income, quarter long-term A fair which This and private recognized In investments in which and investments securities include, adjustment, of quarterly not income to in is fair but to, to subject long-term the may contribute value equity companies future limited net public private equity funds. volatility periods. billion.

to was tax due before expenses RMBX.X profit an in tax. Income year-over-year, billion, increase XXX% increasing primarily

per and net was Our was Baidu was RMBXX.XX. net earnings attributable was net Baidu margin RMBXX.X, under Baidu Net XX%. billion. RMBX.X billion, billion. to were Non-GAAP diluted and for income diluted attributable per ADS and Baidu attributable to Core RMBX.X income Baidu income attributable Non-GAAP was RMBX.X Core to non-GAAP to earnings net ADS RMBX.X income was Core billion, non-GAAP

employees restricted excluding equivalents, RMBXXX were and cash And equivalents, restricted Baidu billion, approximately was IT, RMBX free IT, had was XX, notes [ph] income quarter, billion. in operating RMBX.X cash generated [separate] of as about the Core XXXX. in above RMBX and and cash investments, cash, cash cash billion. cash, and XXXX, short-term XX, flow March On billion. Free [ph] non-GAAP free excluding RMBX.X cash flow. flow, billion March short of billion [iQIYI] cash in were RMBXXX.X As investments XX,XXX

questions. operator, With the now that, call let's open to

Your J.P. [Operator go Yao Alex Instructions] you. Thank first comes from Please ahead. Morgan. with question

is And the strong I trend How how your think ad view my management on verticals? changing the platform? business advertising off-line-related question. vertical the side ads big of then recovery and COVID Congrats how is advertiser growth to is the have about potential lastly, the mid-to-longer-term full-year ad What demand beyond growth for a Thank normalization? you. the for we is on evening Good on thank March? you since for quarter And team. month-on-month a business. the rate? the should what for exposures So, taking

Robin. This Alex. marketing happy see We're progress significant to in is online your for Thank question. Hi, our you business.

online a all revenue year-over-year to declined return revenue that remarkable last comeback high QX, single-digit-growth. Recall year, marketing For made quarters. four for and

I was XX% decline QX, in in QX. X% think and X% QX, it in QX, in X% a

I and by Now, was driven we QX offline the QX. care, in and improving The to just strengthened further – growth has momentum sector of the mentioned are a Particularly, the contributor, revenue prepared in in in a contributes current my a health level. of positive got back verticals during that territory, is remarks. our pre-pandemic the quarter sizable online primarily revenue. to number largest Offline sector our amount been marketing

was we growth revenue the margins The also So, marketing also impact holiday Day positive online really and ecosystem. Labor have mobile our travel of saw is had on apparently, a hot. our really, in the improving,

overall better. the focused a is furthermore, engagement our due in So, The user and and our strong growth to success customer observed And on got Core for margin traffic. experience Baidu user have platform. we

Despite reopening trend. year-on-year and our And And feed demonstrated grew distributed ahead, X% believe this growth. to we also and in search Baidu well-positioned advertisers spending continue MAU App is App content performance-based of to dollars we more benefit more to from the March, allocate Baidu looking offline, for anticipate time people China platforms. robust Baidu our will to

AI a for addition, our opportunity capability generative users complements significant that avenues we and see content. and search create as to drive In access new

into As enhance marketing the apps, further Bot our online we we an integrate for strengthen offering. presenting of to and our search, user opportunity exciting expect our thereby us particularly will up, products, engagement supporting also in ERNIE our commercial traffic ERNIE revenue. the Bot go growth can

and deliver in additional our share business to to With advertisers. revenue. gives sustained us even growth in marketing ERNIE This a Bot ability can its online dimension drive empowered capture confidence we and greater commercial our market further larger value products,

on this we So, the strong our and development marketing revenue online to China's of outperform based GDP. fundamentals, expect growth our

you. Thank

Thank you.

Please Our comes from with go Citigroup. next question ahead. Alicia Yap

Good evening, thank my is related team. to Robin margins spending. and taking Thanks management and for Hi, you. questions My questions.

scale a year, the especially So potential such model, [ph] factors large example, and to as, the Thank impact related – the trend for about rest what language cost-cutting of how large investment we that margin moving the autonomous will operating after the effort [will] last this we be driver others? will major year margin see for the there few given of you. think should core margins,

Alicia. Hello,

me address questions. your Let

we improvements initiatives see to in to number the our margins. transformed you our show and in adapt also but which more it them think of long-term sustainable last or better, operational to growth. have four a all I quarters. the have led drive efficiency, past year of numbers have made implemented notable can our that making not And and overall only which organizations, the These measures will profit leaner agile five

margin operating by just the [indiscernible] has last in number by also also with specifically, result, in More Purely marketing also mentioned, the number together Notably, non-GAAP year [ph] a free to QX. businesses. consider QX year, [form] margins online as this was from a Robin growth entry, mobile from this a XX% ago. in XX% all profit expanded XXX% of is grew as and synergy revenue QX our accompanied one, flow of ecosystem year which over revenue our cash

for and cases, used as the focusing important our that developing more on Baidu significant operating well a carbon profitability Cloud, us. is more for X projects Baidu our on key can Cloud AI and low-margin This as has AI reached Number solutions standard businesses, by is milestone for AI we level. the see

continue enhancing the across And at major the and of particularly the also to in driving. looking margins promising we to areas, continue forward, same our new time, invest ERNIE in while the all operations, Bot aim will intelligent we

actually growth where investments areas reconstructing aligned also significant identified Our we with quite have objectives, long-term a opportunities. our which are two reflect potential

manageable sustainable we I near-term, also to more while [indiscernible]. market in our to aim By the values quite as the we're delivering new growth, large and opportunities try these the we shareholders. investment focusing areas, for on is think shares And drive capture incremental to

same our growth the operational all this long-term on execution the QX, of efficiency disciplined Core the has in of time, these margins Overall, the seize we in Baidu overall and resulted quarters. will future to and the continue try and in focus efforts at opportunities, improvement

you P&L leveraging our the We core will prudently growth. Alicia. further and much, manage so our business improvements and sustainable Thank drive

you. Thank

you. Thank

Our Please go Morgan next with question ahead. from Yu Gary comes Stanley.

Hi, thank related a I question and have you good to evening. cloud.

and us rest So AI expectation from quarters help Cloud could Ernie on coming for cloud adoption? coming revenue of going XQ revenue what's XXXX? And cloud related any growth, the your incremental growth question at the margin. in quantify And growth growth revenue cloud your you in look is on on should revenue margin, normalized for improvement how Thank a margin you. and With and we forward? in

Okay, on Dou. for is cloud. this Thanks the question your

quarter improve So, year, three for the of significant have ways made second since so margins our in consistently we the trying last efforts to Cloud. AI

and Firstly, in are secondly, customers. building resolute across businesses. products solutions cases key we efficiently And then phasing we the them used AI different so are low-quality committed replicate standardized out for and to

we Secondly, to of capabilities continuously in offerings investing their technical cloud reinforce competitiveness. standard the our are

the as a Cloud efforts, of achieved operating quarter. So, a on Baidu non-GAAP profitability the this level AI result

to And China the from meet revenue execute products year, of in we our proposals. In that, will since quarters. now coming continue and and person after our last growth addition clients we in finally it to AI quarter can Cloud's present reopened, is accelerated believe

software education, shares messy in we gain and and cloud the in Moreover, of believe industry. for on. Bot the for have travel Since beta hundreds us from online of the to thousands provide sectors, testing of we ERNIE of ERNIE launch a March, requests enterprise a mid so such as across could its range received customers finance opportunity Bot

overwhelming a the Bot's promising interest debut. start this ERNIE for indicates So,

ERNIE productivity that for to So, and Bot will we are their explore and eager confident their improve our businesses, customers are utilize competitive we their edge. it and so enhance

to on Robin four-layered architecture, We customers just of need the several prepared mentioned as to remarks. our the mentioned, their own cloud has had leverage in Robin large AI as models the now opted advantages Actually, platform. develop

focus sustainable are for business, margins. our cloud still ahead, look we the growth as will improve existing remain on achieving healthy to revenue the striving So, and while we

And the our in so ERNIE sure, the investment the capitalize we opportunity value models. market of so and the we meanwhile, committed Bot. are language huge And large of for will long-term on creation to

we outgrowing we our That's continue long will in in overall, believe revenue it, peers cloud So the run. Gary.

Thank you.

Thank you.

Our Sachs. Kong comes with ahead. from next Goldman question Please Lincoln go

Thank you for management taking question. my

is my Bot. related So, question ERNIE to

So more and basically, can will us line management a reshape and further for how in on in the business our So, growth opportunity model? business especially gain Bot ERNIE elaborate bot Thank monetization should bit about you. in markets? a the ERNIE enterprise the large think help each existing we of share

great has Bot its in Yes, since ERNIE been mid-March. launch making progress

For have to Bot already XX% since its cost upgraded instance, have inference reduce we by release. the and we managed ERNIE Xx, nearly

architecture result While still and of on self-device four improvement cost of our reduction our auto a enhance are layers, model apart our some us AI training enables moved cost. focusing of us which efficiency and competitors for swiftly inference training, a inference. reduce ahead significant This and sets is we've with to

share the strong of position our foundation will upgrade market in marketing, gain enhancing our services, online enables Bot our almost our target using cloud facilitate and services, us markets, strengthen Bot. including and to ERNIE existing products This We're adoption and offerings. This all widespread to smart of devices. and rebuilding ERNIE

ERNIE queuing more is customers the of problems an and of be proving tested particularly of kinds efficiency. to of AI This witnessed generated productivity their have large as and language increasing being business models. Dou of hundreds just and more power [ph] It demonstrates testing. enterprise model. improve by and entities capability, in all recognition AI is enterprises, helping beneficial encouraging Cloud, to Baidu's Foundation] as sectors, launch, AI them and ERNIE thousands [Par and adoption truly Since mentioned, Bot solve In of for is Bot we our

also selective AI into data, Bot, that they and over To provider, developed customers' language a customers using which generative becoming tools customers continuous to large have system are before the training and our assist models refining when improvement use consideration unique of support leads take cloud and better to time. choosing models. more ERNIE we And than technologies capabilities

We our a wanting number of to cloud. growing AI customers see to migrate new

ERNIE anticipate our Bot fleet, we more to and pricing March, of opportunities Cloud. in we we XXX% the sales competitiveness growth our future. saw more power So. in In expect the enhance year-over-year and than AI

industry great cloud presents opportunities the online ERNIE are marketing for Bot significant market, there opportunity while Additionally, adoption, in too.

As I upgrade utilizing promising. consumer family. facing and mentioned received have initial mobile sectors. China's and Baidu engagement, ERNIE applications user in we're The Internet other to user including popular products, competitive Baidu therefore, should and gain implementation, is earlier, Upon reshape traffic Baidu we feedback Search very within App, Bot widespread we landscape the quickly

new avenues growth. We're boundaries of AI progress in exploring pushing of and summary, So, the future. ERNIE Bot's has been encouraging, and very we are about excited the

Thank you.

you. Thank

Miranda America Please go comes next question with from of ahead. Zhuang Securities. Our Bank

competitive launch how management about landscape management long-term, the launching my in on many for And Congratulations near-term question My peers you the to other or is Internet the Thank about competitive questions. taking for to Thank model, With their and results. about language [competition] good large especially [ph]. advantage? the Baidu's the enterprise in think how business? does you. China think facing about

a LLM Yes, constructing this happy rapidly. entry actively ecosystem the stage the to this In very it in and of Moreover, to players that's capabilities the in more to very high we to industry help significant new drive a the we attract AI establish direction. are participants see to divestment, good and very generative a necessary early And question. emerging as few entry. AI are an market. and field Bot develop will ERNIE still the are of companies into innovation our and large possess view, around language barrier present presence the in model

expertise include vast robust and The Also, extensive if appropriate content crucial. substantial requirements amount is Chinese training. to for experience model and in delivering capability AI upfront investment, data Internet mobile users

Given the market, major two probably a highly to end. players winning high three anticipate we such requirements, in concentrated

feedback upgrade large to base, to will which product ERNIE as in products, allows our growth. cutting-edge earlier, are first we revitalize ERNIE of practical able further us lots LLM. service position because address user mentioned approach portfolio in large a mover challenges leading Bot and and and product This China, and and that have sustainable about technology we As features, solutions enhance the user help Bot, into leveraged our to portfolio support entire we get us convert

of has successful distinctive With crucial architecture and four and decrease in AI Baidu quickly role layers we competitive The in improve efficiency this Bot. launch able advantage consisting terms of and a early ERNIE training to gives the developed that ourselves. AI costs we architecture inferencing. in a us architecture, both a of have are capability played

For example, AI training advanced more require effectively including improvements business and the and cloud layer, robust end, and the layer the the this of performance optimized have a Cloud sustain To significant and infrastructure. have and resources, computing own we iteration ongoing our to the optimization work ERNIE this including power, to amount of of most secured GPUs have Bot. and which LLM training efficiency we substantial leading between model substantial framework a for to Developing us enables inferencing, cost PaddlePaddle savings.

build data, our decades in to experience [indiscernible] Chinese in database China's valuable over for enhancing in very knowledge Another language, have critical in operating that us an our and China. two extensive largest is engine including search position point and facts, leadership allowed

ecosystem. of [ph] our And layers any in of and the have [indiscernible], to for vast different all customers experience mobile search offer of combination we're AI on our content We reviewing our amount extensive able stack. [setting] in

can layer have the needs to can those only solution, infrastructure, wants that. offer infrastructure have customer layer, model If needs two. And we the layer, both and we If the we that If if customer the provide framework ready. framework customer we provide that. and needs end-to-end customer

and uniquely advantage, serving positioned competitive are we in cloud market. Based be winner in So, customers. this Thank very will on flexible you. a we confident our this that I'm

very you much. Thank

you. Thank

go Fong Credit Our next question comes from Suisse. Kenneth ahead. Please with

has Hi, noted for draft of good Thank management. on question. generative been Administration China recently AI. taking regulation evening my that you Cyberspace It the a released the

needed investments additional of sort to Thank what the you. So, requirement? are beat

Okay.

to effective Through discussions appropriate search. China's years, search have close principles generative experience online found content. we engaged for applied in XX operating we have quite review During providing with over our AI that communication, for have regulators. we engine similar and are of been Baidu the testing ERNIE has of those in to largest extensive content the Bot,

of also meet compliance development efforts our to requirements. products services ensure both technology and put and a lot We into regulatory applicable

and important hallucinate. For to sensitive we not topics, make AI sure will have

is less or a more final this are trouble requirements all. The not not at yet. LLM task that is probabilistic model, Given

to have it strategy evolves. to update we So, our as continue

China. about informed meet beneficial relevant particularly in solution maintain Furthermore, technology position to of ongoing well AI the tin. supporting an requirements keep the them us and responsible We with share we and insight for This generated the and latest effort dialogue offer actively growth AI. the proactive our authorities generative industry-leading technology,

China. Conversely, regulators do in have challenges content providing companies experience significant generated offering with closely face AI track which not a extensive working appropriate or in will record solutions lacking online of scalable in

has good Baidu that for would long like [leadering] been a time. to emphasize [ph] social AI I for here, So

We of are for in early AI the one of responsive adopters China. use

is it economic to and its industries, best our growth kind China, the in in can and social ERNIE has various We efficiency new productivity increase believe which initiative the contribute China's investment. potential and Bot to across

We energy enhancing resource quality and health improve expect promote care to and in and diagnosis and therefore, disease life. power enterprises innovation agricultural to to Bot to well-being of ERNIE reduction, the allocation, use, industry, people's optimize waste

four That's attach to possibilities and humanity. AI empower XXXX, more Two, freedom four, always should humanity grow Back be to access Also, instead like: the great equal AI then to them. And should and to I and principles AI AI safe with ASICs. AI controllable. we in AI for ASIC. importance should learn up of came bring of humanity. promote replacing technologies Three, One, should

I to want to technology sure the and everyone within co-existence and products, AI to make it understands so comes as that enable these between principles once mankind. AI new Baidu,

we privacy operation and also data In addition, in business prioritize business data and strategy. security

at data we oversee And are establishing AI are for to We a ASICs. for management, dedicated social good. currently committee of promoting another all efforts these and aimed AI have committee

you. Thank

you. Thank

Our Xiong UBS. Wei next question go is ahead. from with Please

is Thank quarter. you and good congrats my for My on Apollo about taking the Go. management questions question

driverless strategy of fully business some your management and operation? for in on the terms the Is updates business could year, robotaxi especially for ride provide plan timeline So, any your hailing there breakeven And the you. could the some this expect of business? Thank we are Apollo you or event year? or scale the Go up what can catalysts that share milestone regarding for

Thanks Wei [indiscernible].

Our public, of the all [world] [ph]. XXX,XXX to we providing this Apollo almost aim hold for autonomous the rides largest Robin by keeps quarter, the growing. the service more autonomous to Go first We driving as mentioned, in continue just and topic. our remains technology to number in ride rides [indiscernible] hailing And expense the provider

business robotaxi cost To operational will on when the in hardware objective, TAM reduce addition, our sustainable In are operation. labor we reach and skill [operations] those achieving model cost, lowering for or focusing gaining it's we are operational costs [ph]. mind efficiency. always of this to and improve and we important costs, building breakeven, a carry make and On the major goal

So, reduce those To we we operation] cost, down crowded the [fully to continue development are to [ph]. working hard to labor of on cut ways find grow expenses.

after As Robin well fully fully of we for closer to provide in goal Beijing, doing In remarks, [indiscernible] Yizhuang in prepared we really the region can adding testing crowded getting mobile are and his the getting [indiscernible]. while [indiscernible]. mentioned our

solid for green reliable roads the because to past. the public is the ride the We purely to This [indiscernible] and open got offer services of the in on mid March. having public of progress light in

users in So and both we last And Wuhan, the since trust gain [ph]. that started government for August. ride services the offering hailing we from a [process]

operations. adding and still are expanding we Now, vehicles

by in keep Go one-third the [indiscernible]. offer will number provided more mentioned rides up this Robin We going As in almost earlier, future. the Apollo of think than Wuhan

and operation. the will provide We safety fully for future driverless reality

the Our continue [indiscernible] decrease. release it we to believe models that hardware after costs, will the and robotics in new of other future,

nationwide. success long-term [ph], a robotaxi. automate in Thank steady achieve the On strategy In to daily are to average future, in continue operational we [operating increase our the housing we will ride efficiency] you.

Thank today. Thank you participating. you. Ladies and gentlemen, that concludes for our for conference

You disconnect lines. may all your