Hype as opposed to fact: How briskly is generative AI replacing the trip trade?

The hype round synthetic intelligence and, extra particularly generative synthetic intelligence, has been palpable over the past just about two years.

Airbnb CEO Brian Chesky is amongst leaders who’ve touted how their firms may evolve because of generative AI. However the trade is but to come back – a reality Chesky said at an tournament he spoke at in September.

“I do not love to make guarantees, and now not ship on them … however, smartly, I am not going to ship at the promise of AI, and nobody will,” stated Chesky. “I feel AI goes to modify the sector a lot more than anyone realizes. I additionally suppose it is going to take means longer than somebody realizes. This decade, issues don’t seem to be going to modify up to other people suppose, and subsequent decade, issues are going to modify much more.”

Have expectancies set by means of trade leaders round AI fallen flat? Is generative AI-powered trade going down as briefly as expected? Those questions really feel related, particularly with the two-year anniversary of ChatGPT’s release coming near subsequent month.

It’s simple to imagine that vary can be speedy, however AI can most effective transfer as rapid as people are ready to transport with it.

Mike Coletta, Phocuswright

Mike Coletta, senior supervisor of study and innovation at Phocuswright, stated there’s a nice dichotomy round AI that makes it tough to wrap one’s head across the era’s evolution.

“At the one hand, it’s by means of a ways the fastest-evolving era in historical past,” Coletta stated. “Then again, people at massive evolve at a collection tempo, each in our figuring out of era and our behaviors with it. So, it’s simple to imagine that vary can be speedy, however AI can most effective transfer as rapid as people are ready to transport with it.”

PhocusWire reached out to trip trade leaders to gauge their evaluations at the tempo at which AI is replacing the trip trade – some imagine Chesky’s level is legitimate, others imagine an AI-powered destiny is right here.

Responses had been won by the use of electronic mail and feature been edited and condensed.

Sundar Narasimhan, senior vice chairman, president of Sabre Labs and product technique, Sabre:

There may be a large number of capital being invested in AI and shortly traders will start to need to see returns for that funding. My view is that – when compared with different rising features like blockchain or VR/AR – AI (particularly LLM / generative AI) is extra tech-ready and embeddable inside of merchandise and use-cases will start to emerge around the trade. Product-market have compatibility may well be tougher to come back by means of, however I feel that we can see extra compelling use situations rising in the following couple of years.

Amy Learn, vice chairman of innovation, Sabre:

AI has been round for some time, and even though the hype has surged for the reason that release of consumer-facing GenAI chatbots, main to large investments within the box, it’s essential to stay a balanced viewpoint. Will we imagine AI is transformative? Completely. However I additionally suppose it’s going to take time to totally notice its possible available in the market. It rings a bell in my memory of the dot-com increase and bust — everybody used to be desperate to dash sooner than we even realized to stroll, and the trade and era wanted time to catch up.

There may be a large number of funding in AI, nevertheless it’s a very powerful to understand that those fashions will most effective be as efficient as the knowledge that drives them. What’s essential now from an trade viewpoint is construction forged foundations—organizing inside sources and knowledge to be future-proof. If we will be able to then follow AI in a significant option to cope with what visitors want by means of simplifying and personalizing their trip reports, one use case at a time, then we can boost up its adoption.

Line Crieloue, vice chairman of symbol and affect, Accor:

We wish to differentiate AI from GenAI. AI has been round for many years now and is already bringing disruptions … like income control or personalization as an example.

GenAI is far more recent. Whilst it’s shifting at a quick tempo, there [are] nonetheless a large number of enhancements but to come back particularly round lowering hallucinations and coaching fashions that carry out in multilingual and multicultural environments. But we do practice already a proliferation of impactful use situations particularly in domain names the place the chance of error has a low affect. This comprises as an example use situations for inside productiveness or assisted gross sales. What is going to needless to say take time is the advance of those fashions and subsequently the acceleration of absolutely automatic use situations comparable to customer-facing options or advertising and marketing similar subjects. When those boost up, it’s going to needless to say disrupt our sector.

Sanjay Mohan, staff leader era officer, MakeMyTrip:

I’ll agree that the GenAI hype has been drummed up fairly aggressively – a little bit too aggressively I’d say. … Fact will ultimately meet up with the hype, and can meet or beat maximum predictions, as I will see. … We’re nonetheless very, very early on this recreation. The prospective is very large, and this GenAI tech is right here for the lengthy haul, so let’s now not pass judgement on too early. It’ll for sure ship towards the guarantees made.

The most obvious use-cases were constructed and shipped. We’ve shipped options that make the most of content material research and summarization in our accommodations funnel. We’ve additionally shipped out some attention-grabbing options using voice in regional Indian languages that successfully help the tip customers whilst they’re in our product flows. That is large for a rustic like India, the place English is proscribed to tier-1 and tier-2 towns most commonly. The GenAI tech is fostering inclusivity, when you view from the language lens. The decision middle brokers are the usage of GenAI-enabled options very successfully. Our workers are the usage of productiveness equipment for coding in addition to research of information every day. So the positive aspects are visual all throughout – externally and internally. The GenAI era is already handing over and developing affect. We, ourselves, are getting formidable in taking over extra complicated options the place GenAI can create important affect in easing the lives of the vacationers.

The era, itself, is evolving at an excessively speedy tempo – on all fronts. The features are getting higher each and every month, the prices are coming down, and the accuracy/latency is getting higher. Should you take a look at the bigger panorama of the tempo of underlying era, and the ambitions of the applying suppliers (together with OTAs), I’m very assured that this decade will finish radically other than what any people may have imagined a couple of years in the past (sooner than GenAI arrived at the scene). The tech is right here to stick, and can ship on all guarantees – most commonly inside of this decade, as I see it. Past that, there’ll [be] a brand new set of desires and imaginations – and that’s for the following decade.

Adam Harris, CEO, CloudBeds:

AI isn’t advancing in our trade as briefly as it might, despite the fact that hospitality used to be one of the vital early adopters of AI, specifically with income control programs, which were in use for many years to optimize pricing and stock. Extra not too long ago, AI chatbots have additionally develop into not unusual, improving visitor reports by means of automating duties like answering FAQs and streamlining bookings.

On the other hand, whilst those level answers are robust, they don’t satisfy the wider promise of AI. Each and every device optimizes only one division — income, visitor products and services, or operations — resulting in remoted positive aspects with out fixing inefficiencies throughout the entire group.

Operating a a hit lodge comes to many interconnected programs, and optimizing only one house whilst leaving others untouched nonetheless leads to inefficiencies. As Invoice Gates has stated, the affect of AI must be as transformative because the introduction of the web or the smartphone —either one of which succeeded as a result of they democratized information and unlocked chances throughout more than one fields. For AI to actually revolutionize hospitality, it wishes to damage down the silos that exist between departments like income, advertising and marketing and visitor revel in.

This fragmentation of information is the primary roadblock. AI can most effective be as efficient as the knowledge it’s fed, and in lots of accommodations, other departments paintings with separate programs that don’t keep up a correspondence with every different. Till an answer is advanced that captures and unifies most of these important information issues, the affect of AI will stay restricted to slim purposes. … Our attached, open platform captures billions of information issues from around the trade, permitting us to construct an advanced Causal AI style that may expect call for and optimize operations throughout all the group. This unified method unlocks the actual possible of AI, serving to accommodations force a complete business technique slightly than optimizing remoted KPIs.

So whilst AI is making strides in hospitality, its transformative possible will most effective be discovered as soon as the trade strikes past departmental silos and adopts platforms in a position to integrating information throughout all the lodge ecosystem. This holistic method is very important to reaching the sweeping adjustments the trade has been looking ahead to.

Alex Bainbridge, CEO, Autoura:

I imagine the trade is right here now, nevertheless it’s exhausting getting AI into the arms of customers because the industry fashions that AI allows do not in reality paintings with how the sectors recently paintings.

For instance, AI commute making plans: Now you might have 10,000 itineraries in a town slightly than 100. How do you commercially function that? And if you’ll be able to’t function them, how do you monetize the ones itineraries past enchantment reserving (which means that it’s important to compete head on with the OTAs on day 1)? 

Timings are extremely tough to pass judgement on right here….. and tech is at all times sluggish [at the industry wide scale] …… however I feel possibly two to 3 years for important trade (US / EU) slightly than finish of decade.

Andres Martinez Artal, founder and CEO, Speakspots.com:

Whilst we now have observed in style integration of “good” chatbots by means of trip startups and incumbents, their obstacles are glaring. They try with ‘proactive’ duties comparable to customizing truly distinctive journeys. Additionally they combat with last-minute trip disruptions because of restricted breadth of real-time information.

To ensure that AI to turn into, slightly than just strengthen the trip making plans and revel in, there are limitations that may take years to conquer. Information silos, comparable to lodge manufacturers, airways and OTA’s personal APIs “by means of invitation most effective” for real-time information, restrict AI brokers’ get entry to to high quality information. The general public web isn’t sufficient.

Higher integration of real-time information from IoT gadgets and broad adoption of voice assistants on wearables may just fortify this. “Anonymized” dialog information is a smart real-time information supply. We additionally want tighter integration throughout reserving, cost, and authentication programs. Accommodations, with direct get entry to to vacationers and, most significantly, recreational suppliers are located to steer AI developments on the vacation spot degree.

Having the traveler’s private AI agent trade knowledge with the lodge’s AI agent will permit accommodations to customise the revel in additional with out want for any receptionist-customer dialog. First, we want large adoption of AI voice assistants by means of society. The problem isn’t for accommodations to undertake AI, lots of them are already actively adopting AI the place they are able to. The hole is cell associated with the acquisition and alternative cycle of cell era. Other folks most effective purchase a brand new cell phone each and every two to 5 years. So even as soon as “new-generation” smartphones are in the marketplace with built-in AI voice era around the board, we’re going to nonetheless be having a look at every other two to 5 years sooner than this tech turns into the mainstream in cell. 

Ilan Twig, CTO and co-founder, Navan:

Whilst the following decade guarantees progressive adjustments pushed by means of generative AI, I am extremely joyful to mention we’re witnessing transformative affects these days. The debut of ChatGPT ignited international pleasure — and sure, some overblown hype. However forward-thinking firms tackling AI’s demanding situations head-on are already reaping outstanding rewards.

…We created Ava, an AI-powered chatbot educated to suppose like a seasoned problem-solver. The adventure wasn’t clean — we confronted hallucinations, biases, and safety demanding situations — however those stumbling blocks fueled our innovation. We advanced Navan Cognition, a game-changing procedure that mixes the most efficient parts of more than a few massive language fashions whilst getting rid of error-prone purposes. The consequences? Impressive. Ava now resolves greater than 50% of reinforce requests, rivaling human brokers in delight ratings and saving us $1 million per 30 days.

Having a look forward, we envision a progressive, proactive UI powered by means of Navan Cognition inside of 5 years. Image intuitive programs that display customers precisely what they want, when they want it. We’re making a bet large on chat changing into the principle interplay mode, with increasingly more subtle voice features built-in into AI bots.

The AI destiny isn’t simply coming — it’s right here, and it’s extra thrilling than we ever imagined. Luck on this AI gold rush calls for diving in early, addressing stumbling blocks, and often pushing obstacles. The transformation is going on now, and it’s only the start of an exciting adventure.

Mike Coletta, senior supervisor of study and innovation, Phocuswright:

Skeptics who level to the joy and sadness curves of applied sciences like digital fact or Web3 must be beneath no illusions about AI … the transformative features of AI are essentially awesome. Thus, those that can transfer quickest, the early adopters, may have a extra distinct benefit than ever. We will be able to see additional step adjustments within the capacity of the underlying fashions in speedy succession, and it’s going to be important for corporations to be ready to undertake them as they come. Whilst there’s at all times possibility in adopting a brand new era, the rewards with AI are transparent.

In trip, those rewards are already being reaped, as we not too long ago detailed in From Buzzword to Backside Line: Conserving Tempo With Gen AI in Trip. And whilst the there’s a lot uncertainty about how AI will affect trade and the sector within the medium to long-term, we’ve defined how trip firms must place themselves for the rewards of the longer term in True Automatic AI in Trip Is Coming, which incorporates the expected affect of self sustaining AI brokers aiding each shoppers with buying groceries and reserving and companies with automation and potency. In essence, apparently most probably that AI will upend the whole lot in trip from consumer interfaces to distribution to the buyer revel in to the body of workers.

Similar to with the web, cell phones and social media, trade appears to be going down slowly for some time after which apparently to occur unexpectedly. Elementary trade will certainly take a few years, as people discover ways to adapt. But it surely’s already overdue to be getting began.

The hype round synthetic intelligence and, extra particularly generative synthetic intelligence, has been palpable over the past just about two years.

Airbnb CEO Brian Chesky is amongst leaders who’ve touted how their firms may evolve because of generative AI. However the trade is but to come back – a reality Chesky said at an tournament he spoke at in September.

“I do not love to make guarantees, and now not ship on them … however, smartly, I am not going to ship at the promise of AI, and nobody will,” stated Chesky. “I feel AI goes to modify the sector a lot more than anyone realizes. I additionally suppose it is going to take means longer than somebody realizes. This decade, issues don’t seem to be going to modify up to other people suppose, and subsequent decade, issues are going to modify much more.”

Have expectancies set by means of trade leaders round AI fallen flat? Is generative AI-powered trade going down as briefly as expected? Those questions really feel related, particularly with the two-year anniversary of ChatGPT’s release coming near subsequent month.

It’s simple to imagine that vary can be speedy, however AI can most effective transfer as rapid as people are ready to transport with it.

Mike Coletta, Phocuswright

Mike Coletta, senior supervisor of study and innovation at Phocuswright, stated there’s a nice dichotomy round AI that makes it tough to wrap one’s head across the era’s evolution.

“At the one hand, it’s by means of a ways the fastest-evolving era in historical past,” Coletta stated. “Then again, people at massive evolve at a collection tempo, each in our figuring out of era and our behaviors with it. So, it’s simple to imagine that vary can be speedy, however AI can most effective transfer as rapid as people are ready to transport with it.”

PhocusWire reached out to trip trade leaders to gauge their evaluations at the tempo at which AI is replacing the trip trade – some imagine Chesky’s level is legitimate, others imagine an AI-powered destiny is right here.

Responses had been won by the use of electronic mail and feature been edited and condensed.

Sundar Narasimhan, senior vice chairman, president of Sabre Labs and product technique, Sabre:

There may be a large number of capital being invested in AI and shortly traders will start to need to see returns for that funding. My view is that – when compared with different rising features like blockchain or VR/AR – AI (particularly LLM / generative AI) is extra tech-ready and embeddable inside of merchandise and use-cases will start to emerge around the trade. Product-market have compatibility may well be tougher to come back by means of, however I feel that we can see extra compelling use situations rising in the following couple of years.

Amy Learn, vice chairman of innovation, Sabre:

AI has been round for some time, and even though the hype has surged for the reason that release of consumer-facing GenAI chatbots, main to large investments within the box, it’s essential to stay a balanced viewpoint. Will we imagine AI is transformative? Completely. However I additionally suppose it’s going to take time to totally notice its possible available in the market. It rings a bell in my memory of the dot-com increase and bust — everybody used to be desperate to dash sooner than we even realized to stroll, and the trade and era wanted time to catch up.

There may be a large number of funding in AI, nevertheless it’s a very powerful to understand that those fashions will most effective be as efficient as the knowledge that drives them. What’s essential now from an trade viewpoint is construction forged foundations—organizing inside sources and knowledge to be future-proof. If we will be able to then follow AI in a significant option to cope with what visitors want by means of simplifying and personalizing their trip reports, one use case at a time, then we can boost up its adoption.

Line Crieloue, vice chairman of symbol and affect, Accor:

We wish to differentiate AI from GenAI. AI has been round for many years now and is already bringing disruptions … like income control or personalization as an example.

GenAI is far more recent. Whilst it’s shifting at a quick tempo, there [are] nonetheless a large number of enhancements but to come back particularly round lowering hallucinations and coaching fashions that carry out in multilingual and multicultural environments. But we do practice already a proliferation of impactful use situations particularly in domain names the place the chance of error has a low affect. This comprises as an example use situations for inside productiveness or assisted gross sales. What is going to needless to say take time is the advance of those fashions and subsequently the acceleration of absolutely automatic use situations comparable to customer-facing options or advertising and marketing similar subjects. When those boost up, it’s going to needless to say disrupt our sector.

Sanjay Mohan, staff leader era officer, MakeMyTrip:

I’ll agree that the GenAI hype has been drummed up fairly aggressively – a little bit too aggressively I’d say. … Fact will ultimately meet up with the hype, and can meet or beat maximum predictions, as I will see. … We’re nonetheless very, very early on this recreation. The prospective is very large, and this GenAI tech is right here for the lengthy haul, so let’s now not pass judgement on too early. It’ll for sure ship towards the guarantees made.

The most obvious use-cases were constructed and shipped. We’ve shipped options that make the most of content material research and summarization in our accommodations funnel. We’ve additionally shipped out some attention-grabbing options using voice in regional Indian languages that successfully help the tip customers whilst they’re in our product flows. That is large for a rustic like India, the place English is proscribed to tier-1 and tier-2 towns most commonly. The GenAI tech is fostering inclusivity, when you view from the language lens. The decision middle brokers are the usage of GenAI-enabled options very successfully. Our workers are the usage of productiveness equipment for coding in addition to research of information every day. So the positive aspects are visual all throughout – externally and internally. The GenAI era is already handing over and developing affect. We, ourselves, are getting formidable in taking over extra complicated options the place GenAI can create important affect in easing the lives of the vacationers.

The era, itself, is evolving at an excessively speedy tempo – on all fronts. The features are getting higher each and every month, the prices are coming down, and the accuracy/latency is getting higher. Should you take a look at the bigger panorama of the tempo of underlying era, and the ambitions of the applying suppliers (together with OTAs), I’m very assured that this decade will finish radically other than what any people may have imagined a couple of years in the past (sooner than GenAI arrived at the scene). The tech is right here to stick, and can ship on all guarantees – most commonly inside of this decade, as I see it. Past that, there’ll [be] a brand new set of desires and imaginations – and that’s for the following decade.

Adam Harris, CEO, CloudBeds:

AI isn’t advancing in our trade as briefly as it might, despite the fact that hospitality used to be one of the vital early adopters of AI, specifically with income control programs, which were in use for many years to optimize pricing and stock. Extra not too long ago, AI chatbots have additionally develop into not unusual, improving visitor reports by means of automating duties like answering FAQs and streamlining bookings.

On the other hand, whilst those level answers are robust, they don’t satisfy the wider promise of AI. Each and every device optimizes only one division — income, visitor products and services, or operations — resulting in remoted positive aspects with out fixing inefficiencies throughout the entire group.

Operating a a hit lodge comes to many interconnected programs, and optimizing only one house whilst leaving others untouched nonetheless leads to inefficiencies. As Invoice Gates has stated, the affect of AI must be as transformative because the introduction of the web or the smartphone —either one of which succeeded as a result of they democratized information and unlocked chances throughout more than one fields. For AI to actually revolutionize hospitality, it wishes to damage down the silos that exist between departments like income, advertising and marketing and visitor revel in.

This fragmentation of information is the primary roadblock. AI can most effective be as efficient as the knowledge it’s fed, and in lots of accommodations, other departments paintings with separate programs that don’t keep up a correspondence with every different. Till an answer is advanced that captures and unifies most of these important information issues, the affect of AI will stay restricted to slim purposes. … Our attached, open platform captures billions of information issues from around the trade, permitting us to construct an advanced Causal AI style that may expect call for and optimize operations throughout all the group. This unified method unlocks the actual possible of AI, serving to accommodations force a complete business technique slightly than optimizing remoted KPIs.

So whilst AI is making strides in hospitality, its transformative possible will most effective be discovered as soon as the trade strikes past departmental silos and adopts platforms in a position to integrating information throughout all the lodge ecosystem. This holistic method is very important to reaching the sweeping adjustments the trade has been looking ahead to.

Alex Bainbridge, CEO, Autoura:

I imagine the trade is right here now, nevertheless it’s exhausting getting AI into the arms of customers because the industry fashions that AI allows do not in reality paintings with how the sectors recently paintings.

For instance, AI commute making plans: Now you might have 10,000 itineraries in a town slightly than 100. How do you commercially function that? And if you’ll be able to’t function them, how do you monetize the ones itineraries past enchantment reserving (which means that it’s important to compete head on with the OTAs on day 1)? 

Timings are extremely tough to pass judgement on right here….. and tech is at all times sluggish [at the industry wide scale] …… however I feel possibly two to 3 years for important trade (US / EU) slightly than finish of decade.

Andres Martinez Artal, founder and CEO, Speakspots.com:

Whilst we now have observed in style integration of “good” chatbots by means of trip startups and incumbents, their obstacles are glaring. They try with ‘proactive’ duties comparable to customizing truly distinctive journeys. Additionally they combat with last-minute trip disruptions because of restricted breadth of real-time information.

To ensure that AI to turn into, slightly than just strengthen the trip making plans and revel in, there are limitations that may take years to conquer. Information silos, comparable to lodge manufacturers, airways and OTA’s personal APIs “by means of invitation most effective” for real-time information, restrict AI brokers’ get entry to to high quality information. The general public web isn’t sufficient.

Higher integration of real-time information from IoT gadgets and broad adoption of voice assistants on wearables may just fortify this. “Anonymized” dialog information is a smart real-time information supply. We additionally want tighter integration throughout reserving, cost, and authentication programs. Accommodations, with direct get entry to to vacationers and, most significantly, recreational suppliers are located to steer AI developments on the vacation spot degree.

Having the traveler’s private AI agent trade knowledge with the lodge’s AI agent will permit accommodations to customise the revel in additional with out want for any receptionist-customer dialog. First, we want large adoption of AI voice assistants by means of society. The problem isn’t for accommodations to undertake AI, lots of them are already actively adopting AI the place they are able to. The hole is cell associated with the acquisition and alternative cycle of cell era. Other folks most effective purchase a brand new cell phone each and every two to 5 years. So even as soon as “new-generation” smartphones are in the marketplace with built-in AI voice era around the board, we’re going to nonetheless be having a look at every other two to 5 years sooner than this tech turns into the mainstream in cell. 

Ilan Twig, CTO and co-founder, Navan:

Whilst the following decade guarantees progressive adjustments pushed by means of generative AI, I am extremely joyful to mention we’re witnessing transformative affects these days. The debut of ChatGPT ignited international pleasure — and sure, some overblown hype. However forward-thinking firms tackling AI’s demanding situations head-on are already reaping outstanding rewards.

…We created Ava, an AI-powered chatbot educated to suppose like a seasoned problem-solver. The adventure wasn’t clean — we confronted hallucinations, biases, and safety demanding situations — however those stumbling blocks fueled our innovation. We advanced Navan Cognition, a game-changing procedure that mixes the most efficient parts of more than a few massive language fashions whilst getting rid of error-prone purposes. The consequences? Impressive. Ava now resolves greater than 50% of reinforce requests, rivaling human brokers in delight ratings and saving us $1 million per 30 days.

Having a look forward, we envision a progressive, proactive UI powered by means of Navan Cognition inside of 5 years. Image intuitive programs that display customers precisely what they want, when they want it. We’re making a bet large on chat changing into the principle interplay mode, with increasingly more subtle voice features built-in into AI bots.

The AI destiny isn’t simply coming — it’s right here, and it’s extra thrilling than we ever imagined. Luck on this AI gold rush calls for diving in early, addressing stumbling blocks, and often pushing obstacles. The transformation is going on now, and it’s only the start of an exciting adventure.

Mike Coletta, senior supervisor of study and innovation, Phocuswright:

Skeptics who level to the joy and sadness curves of applied sciences like digital fact or Web3 must be beneath no illusions about AI … the transformative features of AI are essentially awesome. Thus, those that can transfer quickest, the early adopters, may have a extra distinct benefit than ever. We will be able to see additional step adjustments within the capacity of the underlying fashions in speedy succession, and it’s going to be important for corporations to be ready to undertake them as they come. Whilst there’s at all times possibility in adopting a brand new era, the rewards with AI are transparent.

In trip, those rewards are already being reaped, as we not too long ago detailed in From Buzzword to Backside Line: Conserving Tempo With Gen AI in Trip. And whilst the there’s a lot uncertainty about how AI will affect trade and the sector within the medium to long-term, we’ve defined how trip firms must place themselves for the rewards of the longer term in True Automatic AI in Trip Is Coming, which incorporates the expected affect of self sustaining AI brokers aiding each shoppers with buying groceries and reserving and companies with automation and potency. In essence, apparently most probably that AI will upend the whole lot in trip from consumer interfaces to distribution to the buyer revel in to the body of workers.

Similar to with the web, cell phones and social media, trade appears to be going down slowly for some time after which apparently to occur unexpectedly. Elementary trade will certainly take a few years, as people discover ways to adapt. But it surely’s already overdue to be getting began.