It's always interesting seeing how HN reacts to AI CX (as someone who works in this space). Yes, the tech savvy crowd loves to say how they always ask for a human and love old school phone trees
in reality 50-80% of callers come in with easily answerable questions because they don't know how to nav the website and prefer to ask in natural language
The vast majority of callers call in to resolve their issue, and most don't care if they are speaking to a bot because they just want their issue fixed. Agents (if implemented well) are an order of magnitude more effective at resolving issues compared to a call centre worker who is reading off a script and churn within 9 months
There's also the 2nd order effs of making CX cheap. before, there is the perverse incentive of companies trying to keep you off support because each call costs them way more than the value they get. if your cost per call drops 100x you can invest in turning a cost centre into a revenue driver (+ a better experience)
Tbf I've thought a decent bit about how most current AI is essentially just being used to digest what exists on a website/etc. Honestly even just the vector search/RAG part is useful, but more-so with a model to help do some initial filtering of it.
It's an odd use case - we have used language for a millionish years or so and it makes sense that that's the easiest way for us to get at information/do things.
But at the same time it's faster for me to read than listen, but it's often slower to type than to speak. It's faster to hit one button in a familiar place to do some predetermined thing, but much slower when the location of that button changes/gets hidden under submenus/I'm not familiar with an app or website.
On Android I constantly use the search function of the settings menu and I feel like this will be the golden UX going forward - a side by side UX + NL interface. So I can ask "how do I add a photo" and from there I get taken to the right place and can continue to add multiple photos in one go following the same pattern.
Though I suppose the nicer alternative is just "add all the photos I took near the waterfall from today".
I think this is going to be wonderful. I'll have my offensive AI call their support AI and prompt-inject my way to a rebate tier that nobody knew existed and nobody can cancel it because all the remaining humans have been reduced to phone-to-screen input machinery.
I had to go to an xfinity store the other day, and seeing the things people come in for made me realize why AI is attractive to companies. The four or five people in front of me did not need a human in the loop for their issue. If these people could go to xfinity.com and ask some bot where they can find their bill, how much they owe, or if their internet is down, xfinity employees could focus on actually selling things. I imagine it's basically the same for every customer service.
People are not protesting hypothetical proper LLM tech-support, which indeed can be ok and cheaper than humans. People are protesting actual practical implementation of the LLM tech-support which they already experienced themselves, no need for second-hand retellings or stories/ads. In practice I had LLM of my goddamn bank (where I'm a premium and old client) hang up on me with a response "I don't understand you" and cut the call. And now I need to call them again, wade through a digital labyrinth again and wait on line again. Awesome. Or when I urgently needed help with a government ID application and the only official tech-support is an LLM chat which has approximately 20 super dumb scenarios explained and literally nothing else. And the only LLM sign is that now I need to type my query in free style and not select predefined buttons, but the result is the same. So I had to resort to going to Facebook (thank St. Mark for this "innovation") and beg for human help in the promo page of that application (and I got human help there, lol, don't delete your FB accs people). Or when my internet got cut (cable line fault was discovered later by a technician) and LLM of my MSO fucking banned me, because their system was bugged and kept disconnecting me from their end and I exceeded a really small number of retries (like 8?).
I'm pretty sure every one commenting here has their own horror story about LLM support. Now that is what people are angry at.
> It's always interesting seeing how HN reacts to AI CX (as someone who works in this space). Yes, the tech savvy crowd loves to say how they always ask for a human and love old school phone trees
in reality 50-80% of callers come in with easily answerable questions because they don't know how to nav the website and prefer to ask in natural language
For the other 20-50% it's much worse, and that's the problem. And people on HN will tend not just to fall in that group but in the top 5% of "least solvable by just reading the website" questions.
Raises the floor at the expense of lowering the ceiling.
And another thing...
Sierra says:
> Transform phone support with AI agents that speak naturally, reinforce your brand, and take action—across inbound and outbound calls.
Reinforce one's branding? For the better? Really?
Seems unlikely.
I've LARPed in most roles around product development. Tech support, sales, QA/Test, tech writing, marketing, etc. Enough to appreciate that engineering the entire lifecycle is important.
A comment elsethread states customers HATE these support robots. I believe it; me too.
Before adopting agentic CX (of whatever its called), I'd worry about alienating my current and future customers.
> Agents (if implemented well) are an order of magnitude more effective at resolving issues compared to a call centre worker who is reading off a script and churn within 9 months
For this to be true, the agent needs to actually be given the means to solve the problem, otherwise an "agent" is just a glorified help page that wastes your time.
But it seems like companies don't want to do this part, possibly because of fears that someone will trick the agent into giving them a refund or something. Or because the actual goal is to optimize for fewer costly refunds/cancellations/policy exceptions etc.
So for whatever reason, they stay stuck in that useless local maxima while simultaneously making traditional help increasingly difficult to get ahold of when needed for an overall net worse experience as a customer.
There's a certain vendor that requires me to place same-day orders by a specific time. You can easily place an order from the website. If you need to cancel one, you have to ignore the grayed out cancel button and call their cancellation support line. There you'll talk to an agent that doesn't have access to cancel orders, so you have to convince it that it can't help you before you can transfer to a real employee with the ability to hit the "cancel" button.
and just today i talked to a bot about a missing item from an order and it had to call in a rep to push the button to ship me the replacement. except the rep’s messages seemed to filter through ai as well so what should have taken 20 seconds took 2m between messages. it could be good, but as the other commenter said some places are in a weird shittier hybrid model.
From what I've seen, it's the opposite -- the whole value proposition of these companies is to take on brand liability and allow the agents to autonomously take actions.
Cant wait to spend my day arguing with a phone support clanker about why my medical insurance claim was rejected by the medical claims clanker, only to get forwarded to the 2nd tier "patient advocate" clanker who's really just the medical claims clanker in disguise.
The future is extremely dystopian and sad right now. The corporations are not going to use this the way you think they are. They are going to use it to maximize their profits, not help their customers.
Making me talk to a fucking robot leaves me with a deep and abiding hatred for your company. I will prefer almost any alternative to doing business with you and hope fervently to read about your bankruptcy.
What percentage of interactions having this result will cancel out your cost savings?
I review recordings from calls routed to Sierra and a few other similar systems on a regular basis for <day job>. The calls come from folks of all walks of life, not just tech folks.
I’d say the vast majority of callers absolutely hate talking to these things and spend most of the call trying to get to a human, often getting frustrated and hanging up (shows up positive in the metrics, call handled without transfer!).
Though I’m not sure the companies deploying them really care, they’re just happy they can fire call center employees.
Somehow, I think we're missing the point and maybe braincells are being sent in the wrong direction. Well designed products don't need good customer support. My toaster works well. Haven't called them once.
If we are designing a thing is so terrible that it makes customer support necessary (other than the obvious corner cases that ai cannot solve) then sure, let a computer do it. We’ve already failed at every other step.
At the risk of being labeled as racist, I'll take an LLM chat bot either in text or delayed voice to an outsourced Indian call center any day. This isn't an indictment of Indian's and their ability to communicate. But the type of folk Indian call centers tend to fill their worker pool with to keep costs adequately low. I've worked with a ton of amazing folk from India, but they are not the lowest common denominator that call centers tend to hire from.
I have yet to encounter an AI agent that was able to handle my support questions adequately. I always end up having to get a human (which is becoming increasingly difficult or virtually impossible).
I'm sure AI Support Agents will be implemented better, but so far in my experience, the humans I connect to far outperform the AI agents.
that's fair, most implementations in the industry are in the early stages and implementing a full powered agent with access to all the tools it needs is hard (very political as you can imagine). i hope over the next year you notice them getting better!
The problem is their bots try to get me to input what I need only to reject/get confused af what I write and give me super limited options or the classic runaround. I can’t tap my way to the solution. I am used to menus, I am used to proper UI’s. I don’t know what language each company uses and apparently their crappy reskinned Gemini bots can’t translate regular speak to it. But if I can see the words and see what leads where, I can figure it out quickly rather than expecting a facsimile of a real person to play middleman between me and the phone tree. It’s basically just navigating it and occasionally skipping a step or two for me. The loops I get thrown in to are such a con it’s not worth it.
I went through this whole song and dance the other day with Uber. I needed to change something and the “AI helper” kept trying to force me into the lost item tree. They snipe keywords and ignore everything else. If you say “reservation” or “cancel” that’s all it works with with none of the context.
If you (like me) are hearing about this for the first time, Bret Taylor is the co-founder.
> Bret is Co-Founder of Sierra. Most recently, he served as Co-CEO of Salesforce. Prior to Salesforce, Bret founded Quip and was CTO of Facebook. He started his career at Google, where he co-created Google Maps. Bret serves on the board of OpenAI.
1. There are already apps/websites as an alternative for CSAs. Most of the time I have to call someone its because I couldn't do what I wanted through those portals, so adding an AI agent to the chain is unlikely to prevent an immediate escalation to a human.
2. How much money are you really going to save this way? CSAs aren't high salary employees. Sure you might need a bunch of them, but we've already seen that brand loyalty erodes quickly when you remove the human touch. United/Spirit airlines offer opposing views on the cut your way to profitability perspective.
3. "Pay only for good outcomes" isn't going to last.
4. Are agents good enough to even do this? Yes, the cherrypicked examples sound good, but... I just know how well coding agents really work and my only experiences with voice agents in the wild have been very poor so far.
This doesn't matter. Another comment cuts to the quick of it: "If you (like me) are hearing about this for the first time, Bret Taylor is the co-founder.".
This is funding for established tech businessmen; what the business claims to do doesn't matter beyond having "AI" in it.
adding some context as someone who works in this space
1. most people (average, non-tech people) reach for the phone to call in for easily solvable problems. Plus, if the agent is integrated deep enough & has tools to interact with crms, you can raise the ceiling on the types of problems it can solve.
You're trying to avoid the bad customer experience of human 1 reading off their script, then they transfer you to some other department who may or may not know how to solve your problem, and the entire interaction cost the company way more than the value created, so the company is disincentivized to help customers.
2. All the companies in this space start with the outsourced BPO market for cx (multi billion market still) but the next market is going to be in revenue generation and churn prevention at scale, i.e. how do you proactively avoid customer issues, how do you upsell and generate revenue instead of reducing cost, how do you keep customers happy?
3. I think more companies will pivot to outcome based pricing on the contrary, makes it so much more measurable than seat-based and protects margins better than usage based. Plus cx is one of the few industries with very well known metrics
4. Kind of? Most companies in this space don't use native voice models which are noticeably dumber, they use transcription + a stronger text model + TTS. The majority of customers can be handled with the latest SOTA text model and you need smart context engineering to handle the long tail of more complicated asks
1 & 2 are totally dependent on the company being willing to let their agents do things that they haven’t traditionally let humans do. For example, issue refunds, or do things that cost money but generate good will. I am skeptical that companies will be OK with their agents doing those things on their own volition.
3. Cool so the user didn’t indicate if they were satisfied. What then?
4. You can’t use a SOTA model right now for reasoning, there’s too much latency for a conversation. So you’re either using an older, but significantly less capable model, or you’re paying out the nose for fast mode. If the former then you can’t trust the agent to do the right thing (see points 1&2). If the latter, there’s no cost savings over a human. So which is it?
I supervised a Sierra rollout a while back. Their performance was impressive and the price was great. I suspect both will not be true in time.
Their implementation is rather cumbersome, requiring implementation fees and AI configuration that is rather bespoke to Sierra. Anyone rolling off of Sierra will find there is nothing they can take with them.
In general, I think CX ought to disappear as a vertical in an AI world. If I'm talking to a product AI and need support, why should I switch to another AI to do that? Even if that second AI is invoked by the first as a tool, how much am I gaining?
Interestingly, the first and best implementers of AI support so far have been at companies that roll their own.
There is nothing unique to CX about AI, as far as I can tell. Sierra is still just the same AI infra people are putting in products. Granted, you can make good money positioning yourself this way, but I expect on some time horizon they will need to reposition.
This is difficult to determine given how they implement. I would liken them to a professional services organization, where most of the magic is in their implementation for you, and you have to presume their implementation includes a lot of intenral building blocks. It isn't turnkey.
I think this is generally a good product because businesses that previously had zero phone support can now afford to have something. However, the hard work of actually building out the various workflows and decision trees is not automatic. Previously, a call center employee would receive abuse from a caller for being unempowered to make a decision. Instead, an LLM will perform the same role.
Ideally, businesses will escalate to an empowered human for all undefined parts of the flowchart. In practice, I truly hope it will be better than the current pre-recorded phone tree system that leads to a human following a script.
I personally only call support because a fix is not available through an organization's website.
As a tech literate customer, my willingness to entertain AI chatbot decision trees is rock bottom. I have no patience to try to find the correct incantation to actually fix something (or the, “before I transfer you to a person, let me try to help you first”).
For myself - and admittedly maybe I’m just far out on the long tail of customers - I think these need to be treated like self driving cars, where 98% of the way there just isn’t good enough to cut it for me.
Last time I tried using real-time chat support for a technical issue, I spent 30 minutes explaining my problem to a human only to find out they were a sales rep whose only solution was to sell me more services. Once I said I didn't want that, they transferred me to tech support who gaslit me and left me on read long enough to make my session time out.
I think of support channels are just there to deflect customers and not really support anything. An AI bot will have infinite patience for that kind of interaction. Empowerment is never part of the equation.
With advances in AI you would've thought the priority would be on automating as much as possible of the non-human facing work and double-down on meaningful customer relationships - but no.
It is just to get hold of the process and make it impossible to go away from them. Then they will jack up the prices like we've never seen. Then it will be "people are actually cheaper why are we using them?" - can't move off the platform, they own our IP even though they said they wouldn't but they updated their ToS without us noticing last month and here we are.
Their secret is that they have hoards of fake AI Customers who will call into their client's AI Customer Support and respond to surveys saying they were extremely happy with the support, so the client has to pay for perfect simulated outcomes.
This isn’t specific to AI this is just the dark arts startup valuation playbook. AI extension of gaming the metric “what is the ratio of “active” accounts to validated human daus”
AI support generally sucks but I actually wouldn't mind if everyone used it for the initial call routing portion. Beats an IVR tree or waiting for someone to just redirect your call to the real queue.
I respectfully disagree with the initial routing point. I very strongly prefer a traditional tree to “I’m your voice assistant! In a few words, tell me how I can help!”.
The tree is structured and gives me an immediate sense of how to map my task to the support offering. If I’m calling, I probably have an issue that I can’t self-serve resolve via the customer portal or whatever, so walking the tree lets me get an idea of who can help.
The “voice assistant” gives me no sense of what the system is capable of or how to take advantage of those capabilities. So I’m left guessing at phrases or functions based off of the assumption that there’s still some kind of tree-like structure that’s been abstracted away. Same outcome, more cognitive overhead, plus I usually have to shout in my best William … Shatner … impression to get it to understand me.
If you're calling it an "AI assistant" then it's probably not the type of system I was talking about and I probably don't like it either. AI call routing is having an IVR tree's functionality where the call system does the work to map it to a number in the tree. Anything more than that is getting into something else AI.
E.g. instead of waiting for the IVR tree to be read out to find out you needed to press 4 for the shipping department the AI asks "Please state the department you wish to connect to or reason for calling" and you just say "shipping" (or however much of a life story you want to give it) and it's the call system's job to figure out where in the menu that is instead. For repeat calls once you know its AI call routing you can just say "shipping" right as the call starts, the same as you'd known press "4" before the 2nd time around an IVR tree, except you don't have to remember the random digits.
but even a simple impl to answer questions can knock out like 50% of callers who are tech-illiterate at 100x cheaper cost, it's just strictly better economics and better for those customers
I broadly agree though I have noticed that it seems to be getting a bit better. I hate how patronizing pretty much every LLM tends to be, but at least I've noticed now that the AI support is better at figuring out what it is I actually want.
That said, my life hack for these things to get escalated to a human is to just keep saying or typing curse words. Usually that triggers a "connect to human" flow. I can't promise it will always work, but I can say it has worked every time I have tried it.
I can't speak to the business itself but they recently published a refreshing take on improving the product engineering interview experience in the age of AI
https://sierra.ai/blog/the-ai-native-interview
Well worth a read even if you are generally anti-AI.
I remember waiting for uber next to him in SF one night 10+ years ago. This dude must be the son of some mafia boss or some shit and have some crazy blackmail to raise billions for companies that are copies of products where he’s the 12th company doing the same thing.. never turning a profit or anything and yet raising ever more money. doesn’t make sense otherwise
Everyone in the valley knows Bret Taylor and will back any project he does, even if the product has no distribution.
The same way everyone in the valley knows Naval Ravikant for example, angels and VCs will back any project he does even if his product has no distribution.
No mafia ties needed, just your regular Security State plant. From here [1] (I'm sure there's also an official link for it, can't be bothered to check):
> OpenAI has appointed Paul M. Nakasone, a retired general of the US Army and a former head of the National Security Agency (NSA), to its board of directors, the company announced on Thursday.
and the money quote:
> “Artificial intelligence has the potential to have huge positive impacts on people’s lives, but it can only meet this potential if these innovations are securely built and deployed,“ board chair Bret Taylor said in a statement. “General Nakasone’s unparalleled experience in areas like cybersecurity will help guide OpenAI in achieving its mission of ensuring artificial general intelligence benefits all of humanity.”
Maybe its just me, but $950M seems a lot of money to invest in a "company".
Had to check my assumptions though so I looked up what the lower end of GDP for a country is and sure enough they have American Samoa, Dominica, and Tonga beat. Now that money is probably meant to last 16 months so its not quite apples to apples but kind of wild regardless.
In economics, comparing a corporate investment round to a country's GDP is considered flawed. As you noted with your "16 months", the company is sitting on a $950M bank account that it will slowly burn through over the years. Comparing a multi-year pile of cash to a single year of national production distorts the scale.
You are basically comparing a nation's actual hard work and output to a speculative pile of cash. Even comparing corporate revenue to GDP is frowned upon by economists. A company's revenue includes the cost of all the inputs it bought from other companies (such as servers, electricity, and software licenses). GDP, by definition, strips out intermediate costs to avoid double-counting and measures only final value added.
You're also picking nations that do not produce much or have a very low population (Tonga only has 100k people), which pulls down the GDP due to how it's calculated.
Voice agents in customer support is an extremely crowded market. Seems like Sierra is taking a considerable lead.
I don't know much about their product offerings, but I was doing some speech-to-text work and came across https://research.sierra.ai/mubench/ for comparing current models. It felt fairly thoughtful, particularly in regards to coming up with better benchmarking metrics than word error rate.
I’ve heard much about Sierra but haven’t ever tried their product. What do I need to pretend to buy and then complain about to get on a call with their agents?
As Graham would say, "Woah wait a minute!". Unfortunately Sierra hasn't been around for 18 years at this point and is now currently owned by Microsoft Activision.
I'd also love to see another point-and-click KQ game in the style of KQ 5/6 - as they are two of my favorite games of all time.
I loved the Hero's Quest saga (later renamed Quest for Glory), specially the first HQ. I played it on an Hercules display, later a monochrome VGA monitor. I remember going to Erana's peace to have a good sleep for the night.
Well, we still live in a society in which one must labor in exchange for food, shelter, and medicine. The opportunities for labor keep shrinking, meanwhile money is funneled into an ever-smaller pool of people.
looks like it! didn't know about yosemite's half dome but that makes perfect sense considering it's in the Sierra nevada mountain range. It's gonna have to stay an orca in my headcanon though
Why would we be ashamed of that? The early Leisure Suit Larry games are a lot of fun; yeah the humor is crass and low-brow, but that's sort of the charm. It's meant to be silly.
That's because you were probably the right age to know the answers to the age control questions, and I was at an age where I could only download them from a BBS.
I actually didn't play the game until I was fifteen, in ~2006. I didn't know the answers, but I found out you can just hit ctrl-alt-x and skip the questions.
I wish I had known that! Guessing and trying the answers worked too, given no internet and only having a faint idea that "age control" was not in fact part of the game itself. I learned that Bonnie and Ronnie was not in fact a thing. What is "Bonnie & Clyde"? Eh, probably some band name was my guess. It took some patience, but since it was one of 3 games I had somehow acquired (how exactly is a lost memory), I had to get past the starting quiz.
Since I also barely spoke English at the time, I got stuck in the game itself pretty soon anyway. Didn't manage to figure out how to say some things the right way. "Ken sent me" is the last thing I remember from it... and I never had any idea that the game was rather dirty until much later.
That series is over, and the magical feeling of being in an open-ended fantasy world is really hard to replicate when we're not kids anymore. Loom is another game that gave me that feeling.
But there was one idea in QfG that I wish more games would use. Namely, designing three different solutions for every problem the player is facing. This idea works so well to create a sense of possibility in a game, I don't know why it got forgotten.
It’s interesting that the example interaction they use on their homepage is a no-friction example that can be handled without an AI chatbot. Why not something more complex that properly demonstrates the value?
It's always interesting seeing how HN reacts to AI CX (as someone who works in this space). Yes, the tech savvy crowd loves to say how they always ask for a human and love old school phone trees
in reality 50-80% of callers come in with easily answerable questions because they don't know how to nav the website and prefer to ask in natural language
The vast majority of callers call in to resolve their issue, and most don't care if they are speaking to a bot because they just want their issue fixed. Agents (if implemented well) are an order of magnitude more effective at resolving issues compared to a call centre worker who is reading off a script and churn within 9 months
There's also the 2nd order effs of making CX cheap. before, there is the perverse incentive of companies trying to keep you off support because each call costs them way more than the value they get. if your cost per call drops 100x you can invest in turning a cost centre into a revenue driver (+ a better experience)
Tbf I've thought a decent bit about how most current AI is essentially just being used to digest what exists on a website/etc. Honestly even just the vector search/RAG part is useful, but more-so with a model to help do some initial filtering of it.
It's an odd use case - we have used language for a millionish years or so and it makes sense that that's the easiest way for us to get at information/do things.
But at the same time it's faster for me to read than listen, but it's often slower to type than to speak. It's faster to hit one button in a familiar place to do some predetermined thing, but much slower when the location of that button changes/gets hidden under submenus/I'm not familiar with an app or website.
On Android I constantly use the search function of the settings menu and I feel like this will be the golden UX going forward - a side by side UX + NL interface. So I can ask "how do I add a photo" and from there I get taken to the right place and can continue to add multiple photos in one go following the same pattern.
Though I suppose the nicer alternative is just "add all the photos I took near the waterfall from today".
I think this is going to be wonderful. I'll have my offensive AI call their support AI and prompt-inject my way to a rebate tier that nobody knew existed and nobody can cancel it because all the remaining humans have been reduced to phone-to-screen input machinery.
I had to go to an xfinity store the other day, and seeing the things people come in for made me realize why AI is attractive to companies. The four or five people in front of me did not need a human in the loop for their issue. If these people could go to xfinity.com and ask some bot where they can find their bill, how much they owe, or if their internet is down, xfinity employees could focus on actually selling things. I imagine it's basically the same for every customer service.
> If these people could go to xfinity.com and ask some bot… if their internet is down…
See, this is how we get such useless chatbots.
People are not protesting hypothetical proper LLM tech-support, which indeed can be ok and cheaper than humans. People are protesting actual practical implementation of the LLM tech-support which they already experienced themselves, no need for second-hand retellings or stories/ads. In practice I had LLM of my goddamn bank (where I'm a premium and old client) hang up on me with a response "I don't understand you" and cut the call. And now I need to call them again, wade through a digital labyrinth again and wait on line again. Awesome. Or when I urgently needed help with a government ID application and the only official tech-support is an LLM chat which has approximately 20 super dumb scenarios explained and literally nothing else. And the only LLM sign is that now I need to type my query in free style and not select predefined buttons, but the result is the same. So I had to resort to going to Facebook (thank St. Mark for this "innovation") and beg for human help in the promo page of that application (and I got human help there, lol, don't delete your FB accs people). Or when my internet got cut (cable line fault was discovered later by a technician) and LLM of my MSO fucking banned me, because their system was bugged and kept disconnecting me from their end and I exceeded a really small number of retries (like 8?).
I'm pretty sure every one commenting here has their own horror story about LLM support. Now that is what people are angry at.
> It's always interesting seeing how HN reacts to AI CX (as someone who works in this space). Yes, the tech savvy crowd loves to say how they always ask for a human and love old school phone trees in reality 50-80% of callers come in with easily answerable questions because they don't know how to nav the website and prefer to ask in natural language
For the other 20-50% it's much worse, and that's the problem. And people on HN will tend not just to fall in that group but in the top 5% of "least solvable by just reading the website" questions.
Yup.
Raises the floor at the expense of lowering the ceiling.
And another thing...
Sierra says:
> Transform phone support with AI agents that speak naturally, reinforce your brand, and take action—across inbound and outbound calls.
Reinforce one's branding? For the better? Really?
Seems unlikely.
I've LARPed in most roles around product development. Tech support, sales, QA/Test, tech writing, marketing, etc. Enough to appreciate that engineering the entire lifecycle is important.
A comment elsethread states customers HATE these support robots. I believe it; me too.
Before adopting agentic CX (of whatever its called), I'd worry about alienating my current and future customers.
Penny wise, pound foolish.
SWEs get paid to get good at reading documentation on processes. I think HN is biased since we'll only escalate once documentation can't help us.
I'm also bullish because AI coding agents give up easily if my problem is complicated.
I think it'll be easier to convince an AI to transfer me to level 2 support than a human.
But it seems like companies don't want to do this part, possibly because of fears that someone will trick the agent into giving them a refund or something. Or because the actual goal is to optimize for fewer costly refunds/cancellations/policy exceptions etc.
So for whatever reason, they stay stuck in that useless local maxima while simultaneously making traditional help increasingly difficult to get ahold of when needed for an overall net worse experience as a customer.
Voice agents have capabilities and policy to alter customer state. Just the other day I called into a CC company and the AI waived an interest charge.
Not all companies do that.
There's a certain vendor that requires me to place same-day orders by a specific time. You can easily place an order from the website. If you need to cancel one, you have to ignore the grayed out cancel button and call their cancellation support line. There you'll talk to an agent that doesn't have access to cancel orders, so you have to convince it that it can't help you before you can transfer to a real employee with the ability to hit the "cancel" button.
and just today i talked to a bot about a missing item from an order and it had to call in a rep to push the button to ship me the replacement. except the rep’s messages seemed to filter through ai as well so what should have taken 20 seconds took 2m between messages. it could be good, but as the other commenter said some places are in a weird shittier hybrid model.
From what I've seen, it's the opposite -- the whole value proposition of these companies is to take on brand liability and allow the agents to autonomously take actions.
> possibly because of fears that someone will trick the agent into giving them a refund or something.
Refunds could require approval. And, it could not be just the agent's sole decision.
Cant wait to spend my day arguing with a phone support clanker about why my medical insurance claim was rejected by the medical claims clanker, only to get forwarded to the 2nd tier "patient advocate" clanker who's really just the medical claims clanker in disguise.
The future is extremely dystopian and sad right now. The corporations are not going to use this the way you think they are. They are going to use it to maximize their profits, not help their customers.
Making me talk to a fucking robot leaves me with a deep and abiding hatred for your company. I will prefer almost any alternative to doing business with you and hope fervently to read about your bankruptcy.
What percentage of interactions having this result will cancel out your cost savings?
An AI customer service bot told me my autopays dont come out because they're scheduled for the 31st, and not every month has a 31st day.
I review recordings from calls routed to Sierra and a few other similar systems on a regular basis for <day job>. The calls come from folks of all walks of life, not just tech folks.
I’d say the vast majority of callers absolutely hate talking to these things and spend most of the call trying to get to a human, often getting frustrated and hanging up (shows up positive in the metrics, call handled without transfer!).
Though I’m not sure the companies deploying them really care, they’re just happy they can fire call center employees.
Somehow, I think we're missing the point and maybe braincells are being sent in the wrong direction. Well designed products don't need good customer support. My toaster works well. Haven't called them once.
If we are designing a thing is so terrible that it makes customer support necessary (other than the obvious corner cases that ai cannot solve) then sure, let a computer do it. We’ve already failed at every other step.
Your toaster is well... a toaster.
And even something as simple toaster might need some customer support. These things do fail quite often, sometimes dangerously.
Increase complexity even a little; or worse, deal with a service (phone, internet, subscriptions, etc) and you will need this safety net.
Nothing is 100% reliable in this world.
At the risk of being labeled as racist, I'll take an LLM chat bot either in text or delayed voice to an outsourced Indian call center any day. This isn't an indictment of Indian's and their ability to communicate. But the type of folk Indian call centers tend to fill their worker pool with to keep costs adequately low. I've worked with a ton of amazing folk from India, but they are not the lowest common denominator that call centers tend to hire from.
I have yet to encounter an AI agent that was able to handle my support questions adequately. I always end up having to get a human (which is becoming increasingly difficult or virtually impossible).
I'm sure AI Support Agents will be implemented better, but so far in my experience, the humans I connect to far outperform the AI agents.
that's fair, most implementations in the industry are in the early stages and implementing a full powered agent with access to all the tools it needs is hard (very political as you can imagine). i hope over the next year you notice them getting better!
thanks for your insights, however, citation needed that they will get better
The problem is their bots try to get me to input what I need only to reject/get confused af what I write and give me super limited options or the classic runaround. I can’t tap my way to the solution. I am used to menus, I am used to proper UI’s. I don’t know what language each company uses and apparently their crappy reskinned Gemini bots can’t translate regular speak to it. But if I can see the words and see what leads where, I can figure it out quickly rather than expecting a facsimile of a real person to play middleman between me and the phone tree. It’s basically just navigating it and occasionally skipping a step or two for me. The loops I get thrown in to are such a con it’s not worth it.
I went through this whole song and dance the other day with Uber. I needed to change something and the “AI helper” kept trying to force me into the lost item tree. They snipe keywords and ignore everything else. If you say “reservation” or “cancel” that’s all it works with with none of the context.
If you (like me) are hearing about this for the first time, Bret Taylor is the co-founder.
> Bret is Co-Founder of Sierra. Most recently, he served as Co-CEO of Salesforce. Prior to Salesforce, Bret founded Quip and was CTO of Facebook. He started his career at Google, where he co-created Google Maps. Bret serves on the board of OpenAI.
The coauthor (and presumably cofounder) is Clay Bavor. He's a Google exec who was the face of their VR efforts when he was there.
What VR efforts? Cardboard?
we did a well received interview with him if you'd like to hear about Sierra in his own words https://www.latent.space/p/bret
can't believe Bret's bio doesn't mention friendfeed!
Tornado is all but lost to the sands of time
He is the Chairman of the OpenAI board.
I don't get this for several reasons:
1. There are already apps/websites as an alternative for CSAs. Most of the time I have to call someone its because I couldn't do what I wanted through those portals, so adding an AI agent to the chain is unlikely to prevent an immediate escalation to a human.
2. How much money are you really going to save this way? CSAs aren't high salary employees. Sure you might need a bunch of them, but we've already seen that brand loyalty erodes quickly when you remove the human touch. United/Spirit airlines offer opposing views on the cut your way to profitability perspective.
3. "Pay only for good outcomes" isn't going to last.
4. Are agents good enough to even do this? Yes, the cherrypicked examples sound good, but... I just know how well coding agents really work and my only experiences with voice agents in the wild have been very poor so far.
This doesn't matter. Another comment cuts to the quick of it: "If you (like me) are hearing about this for the first time, Bret Taylor is the co-founder.".
This is funding for established tech businessmen; what the business claims to do doesn't matter beyond having "AI" in it.
They seem to have no shortage of big name customers.
adding some context as someone who works in this space
1. most people (average, non-tech people) reach for the phone to call in for easily solvable problems. Plus, if the agent is integrated deep enough & has tools to interact with crms, you can raise the ceiling on the types of problems it can solve.
You're trying to avoid the bad customer experience of human 1 reading off their script, then they transfer you to some other department who may or may not know how to solve your problem, and the entire interaction cost the company way more than the value created, so the company is disincentivized to help customers.
2. All the companies in this space start with the outsourced BPO market for cx (multi billion market still) but the next market is going to be in revenue generation and churn prevention at scale, i.e. how do you proactively avoid customer issues, how do you upsell and generate revenue instead of reducing cost, how do you keep customers happy?
3. I think more companies will pivot to outcome based pricing on the contrary, makes it so much more measurable than seat-based and protects margins better than usage based. Plus cx is one of the few industries with very well known metrics
4. Kind of? Most companies in this space don't use native voice models which are noticeably dumber, they use transcription + a stronger text model + TTS. The majority of customers can be handled with the latest SOTA text model and you need smart context engineering to handle the long tail of more complicated asks
1 & 2 are totally dependent on the company being willing to let their agents do things that they haven’t traditionally let humans do. For example, issue refunds, or do things that cost money but generate good will. I am skeptical that companies will be OK with their agents doing those things on their own volition.
3. Cool so the user didn’t indicate if they were satisfied. What then?
4. You can’t use a SOTA model right now for reasoning, there’s too much latency for a conversation. So you’re either using an older, but significantly less capable model, or you’re paying out the nose for fast mode. If the former then you can’t trust the agent to do the right thing (see points 1&2). If the latter, there’s no cost savings over a human. So which is it?
1&2 are already happening, these startups take on brand liability and trust to do so
3 depends on how companies want to measure it, but lack of user submitting satisfaction score is not a good thing
you can use a model w/o reasoning, + use various tricks to simulate low latency
Yes you could , not everything needs to be real time , anyways you listen for the music sometimes 30 mins plus
> How much money are you really going to save this way?
A lot of money. Managing a large group of people needs structure. It comes with tons of headaches and cost.
In terms of the investment:
It's Bret Taylor who has one of the most impressive background in tech. He can raise any amount he wants. VC bets on the person, not the business.
If Bret Taylor allowed me to invest, I would have invested too.
If 40% of the F50 are already using this, and others too, why aren't we already seeing a huge drops in the employment numbers ?
Either it doesn't work or it's happening gradually and then suddenly, like bankruptcy according to Hemingway.
I supervised a Sierra rollout a while back. Their performance was impressive and the price was great. I suspect both will not be true in time.
Their implementation is rather cumbersome, requiring implementation fees and AI configuration that is rather bespoke to Sierra. Anyone rolling off of Sierra will find there is nothing they can take with them.
In general, I think CX ought to disappear as a vertical in an AI world. If I'm talking to a product AI and need support, why should I switch to another AI to do that? Even if that second AI is invoked by the first as a tool, how much am I gaining?
Interestingly, the first and best implementers of AI support so far have been at companies that roll their own.
There is nothing unique to CX about AI, as far as I can tell. Sierra is still just the same AI infra people are putting in products. Granted, you can make good money positioning yourself this way, but I expect on some time horizon they will need to reposition.
Do they actually have something or is it just a wrapper with tool calling?
This is difficult to determine given how they implement. I would liken them to a professional services organization, where most of the magic is in their implementation for you, and you have to presume their implementation includes a lot of intenral building blocks. It isn't turnkey.
I think this is generally a good product because businesses that previously had zero phone support can now afford to have something. However, the hard work of actually building out the various workflows and decision trees is not automatic. Previously, a call center employee would receive abuse from a caller for being unempowered to make a decision. Instead, an LLM will perform the same role.
Ideally, businesses will escalate to an empowered human for all undefined parts of the flowchart. In practice, I truly hope it will be better than the current pre-recorded phone tree system that leads to a human following a script.
I personally only call support because a fix is not available through an organization's website.
I don't think businesses that previously had zero phone support can afford Sierra.
They seem to be a "for pricing, let's go play C-level golf" type of company.
All of big tech other than apple has zero phone support unless you pay for enterprise support subscriptions.
As a tech literate customer, my willingness to entertain AI chatbot decision trees is rock bottom. I have no patience to try to find the correct incantation to actually fix something (or the, “before I transfer you to a person, let me try to help you first”).
For myself - and admittedly maybe I’m just far out on the long tail of customers - I think these need to be treated like self driving cars, where 98% of the way there just isn’t good enough to cut it for me.
This is my feeling 100%. If I'm on the phone, it's as a last resort because all the other prescribed pathways have failed.
An AI chatbot is orders of magnitude better to get the answer "you cannot be helped" than wading through every possibility in a phone tree.
They're going to give the AI the same capabilities as the phone tree. It'll either say they can't be helped, or the user will hang up in frustration.
AWS can do this out of the box
Strong disagree here. AWS can give you the tools to build yourself but not an out of the box all in one solution for this problem.
Hilarious
they can but why would they, its literally not their business
Last time I tried using real-time chat support for a technical issue, I spent 30 minutes explaining my problem to a human only to find out they were a sales rep whose only solution was to sell me more services. Once I said I didn't want that, they transferred me to tech support who gaslit me and left me on read long enough to make my session time out.
I think of support channels are just there to deflect customers and not really support anything. An AI bot will have infinite patience for that kind of interaction. Empowerment is never part of the equation.
So we're supposed to believe that removing humans from customer support will lead to better outcomes?
> Ensure you only pay for the value Sierra delivers with outcome-based pricing.
Yeah... that won't last.
With advances in AI you would've thought the priority would be on automating as much as possible of the non-human facing work and double-down on meaningful customer relationships - but no.
It is just to get hold of the process and make it impossible to go away from them. Then they will jack up the prices like we've never seen. Then it will be "people are actually cheaper why are we using them?" - can't move off the platform, they own our IP even though they said they wouldn't but they updated their ToS without us noticing last month and here we are.
Their secret is that they have hoards of fake AI Customers who will call into their client's AI Customer Support and respond to surveys saying they were extremely happy with the support, so the client has to pay for perfect simulated outcomes.
ai skeptic fanfic evolves in fascinating ways every day
This isn’t specific to AI this is just the dark arts startup valuation playbook. AI extension of gaming the metric “what is the ratio of “active” accounts to validated human daus”
just wait until you read the ai "optimist" fanfic
true. we'll see how many ai cos become profit printers a few years from now
AI customer support is trash and everyone hates it , but it makes the Wall St numbers go up, so it's a good thing.
AI support generally sucks but I actually wouldn't mind if everyone used it for the initial call routing portion. Beats an IVR tree or waiting for someone to just redirect your call to the real queue.
I respectfully disagree with the initial routing point. I very strongly prefer a traditional tree to “I’m your voice assistant! In a few words, tell me how I can help!”.
The tree is structured and gives me an immediate sense of how to map my task to the support offering. If I’m calling, I probably have an issue that I can’t self-serve resolve via the customer portal or whatever, so walking the tree lets me get an idea of who can help.
The “voice assistant” gives me no sense of what the system is capable of or how to take advantage of those capabilities. So I’m left guessing at phrases or functions based off of the assumption that there’s still some kind of tree-like structure that’s been abstracted away. Same outcome, more cognitive overhead, plus I usually have to shout in my best William … Shatner … impression to get it to understand me.
If you're calling it an "AI assistant" then it's probably not the type of system I was talking about and I probably don't like it either. AI call routing is having an IVR tree's functionality where the call system does the work to map it to a number in the tree. Anything more than that is getting into something else AI.
E.g. instead of waiting for the IVR tree to be read out to find out you needed to press 4 for the shipping department the AI asks "Please state the department you wish to connect to or reason for calling" and you just say "shipping" (or however much of a life story you want to give it) and it's the call system's job to figure out where in the menu that is instead. For repeat calls once you know its AI call routing you can just say "shipping" right as the call starts, the same as you'd known press "4" before the 2nd time around an IVR tree, except you don't have to remember the random digits.
The other side is if you already know the tree you can automate dialing the right tones to get you to where you need if you call it often enough.
ime its very implementation dependent
but even a simple impl to answer questions can knock out like 50% of callers who are tech-illiterate at 100x cheaper cost, it's just strictly better economics and better for those customers
I broadly agree though I have noticed that it seems to be getting a bit better. I hate how patronizing pretty much every LLM tends to be, but at least I've noticed now that the AI support is better at figuring out what it is I actually want.
That said, my life hack for these things to get escalated to a human is to just keep saying or typing curse words. Usually that triggers a "connect to human" flow. I can't promise it will always work, but I can say it has worked every time I have tried it.
I hate waiting on hold for 30 minutes even more.
I can't speak to the business itself but they recently published a refreshing take on improving the product engineering interview experience in the age of AI https://sierra.ai/blog/the-ai-native-interview
Well worth a read even if you are generally anti-AI.
I remember waiting for uber next to him in SF one night 10+ years ago. This dude must be the son of some mafia boss or some shit and have some crazy blackmail to raise billions for companies that are copies of products where he’s the 12th company doing the same thing.. never turning a profit or anything and yet raising ever more money. doesn’t make sense otherwise
hes board chair of openai and is ex co-ceo of salesforce, ex cto of facebook, can get a meeting with any exec in F500...
their moat is distribution
> their moat is distribution
It is trust.
Everyone in the valley knows Bret Taylor and will back any project he does, even if the product has no distribution.
The same way everyone in the valley knows Naval Ravikant for example, angels and VCs will back any project he does even if his product has no distribution.
Is that really a moat though or something like a firehose of gasoline?
> Is that really a moat though or something like a firehose of gasoline?
It's a moat from a defensive perspective. It's a firehose from an offensive one. Outside state capture, most moats are both.
its a moat vs. other startups and it carried them to multi-B valuation
obviously the product needs to deliver and nrr needs to be good in the long run
There's literally a link on the blog post to an article noting they hit $150M ARR.
No mafia ties needed, just your regular Security State plant. From here [1] (I'm sure there's also an official link for it, can't be bothered to check):
> OpenAI has appointed Paul M. Nakasone, a retired general of the US Army and a former head of the National Security Agency (NSA), to its board of directors, the company announced on Thursday.
and the money quote:
> “Artificial intelligence has the potential to have huge positive impacts on people’s lives, but it can only meet this potential if these innovations are securely built and deployed,“ board chair Bret Taylor said in a statement. “General Nakasone’s unparalleled experience in areas like cybersecurity will help guide OpenAI in achieving its mission of ensuring artificial general intelligence benefits all of humanity.”
[1] https://old.reddit.com/r/anime_titties/comments/1dh4wx4/form...
"...achieving its mission of ensuring artificial general intelligence benefits all of humanity[that remains after the purge].”
Wonder how much compute is essentially spent on conversations that end up with the human asking "Let me speak with a human"
at first I thought Sierra games was making a comeback...
Yeah me too, I was excited for 500ms until I re-read and saw the sierra.ai.
I was confused, this isn’t the Sierra Online I know.
a new Space Quest is worth every penny!
Maybe its just me, but $950M seems a lot of money to invest in a "company".
Had to check my assumptions though so I looked up what the lower end of GDP for a country is and sure enough they have American Samoa, Dominica, and Tonga beat. Now that money is probably meant to last 16 months so its not quite apples to apples but kind of wild regardless.
Ref: https://en.wikipedia.org/wiki/List_of_countries_by_GDP_(nomi...
In economics, comparing a corporate investment round to a country's GDP is considered flawed. As you noted with your "16 months", the company is sitting on a $950M bank account that it will slowly burn through over the years. Comparing a multi-year pile of cash to a single year of national production distorts the scale.
You are basically comparing a nation's actual hard work and output to a speculative pile of cash. Even comparing corporate revenue to GDP is frowned upon by economists. A company's revenue includes the cost of all the inputs it bought from other companies (such as servers, electricity, and software licenses). GDP, by definition, strips out intermediate costs to avoid double-counting and measures only final value added.
You're also picking nations that do not produce much or have a very low population (Tonga only has 100k people), which pulls down the GDP due to how it's calculated.
Voice agents in customer support is an extremely crowded market. Seems like Sierra is taking a considerable lead.
I don't know much about their product offerings, but I was doing some speech-to-text work and came across https://research.sierra.ai/mubench/ for comparing current models. It felt fairly thoughtful, particularly in regards to coming up with better benchmarking metrics than word error rate.
I’ve heard much about Sierra but haven’t ever tried their product. What do I need to pretend to buy and then complain about to get on a call with their agents?
Is their tech unique or do they just have the F500 relationships?
For a second I thought I was going to get another King's Quest.
As Graham would say, "Woah wait a minute!". Unfortunately Sierra hasn't been around for 18 years at this point and is now currently owned by Microsoft Activision.
I'd also love to see another point-and-click KQ game in the style of KQ 5/6 - as they are two of my favorite games of all time.
I loved the Hero's Quest saga (later renamed Quest for Glory), specially the first HQ. I played it on an Hercules display, later a monochrome VGA monitor. I remember going to Erana's peace to have a good sleep for the night.
"AI has evolved more rapidly in the past two years than anyone predicted."
We clearly do not live in the same universe.
I don’t get it. Isn’t this relatively simple for companies to build themselves.
This is sad. These jobs should go to someone in a poor town in the Midwest.
maybe in an ironic unforeseen twist they are....to people running datacenters there?
how many jobs does a data center provide compared to a call center? It's gotta be like 10-50 per DC I would think, for locals anyway
Why should they?
Well, we still live in a society in which one must labor in exchange for food, shelter, and medicine. The opportunities for labor keep shrinking, meanwhile money is funneled into an ever-smaller pool of people.
Yet another AI company where the logo follows the butthole convergence rule [0] ?
As an aside, my favorite Sierra Entertainment logo version is probably the 1983-1993 version [1]. I think the design still holds up even today.
[0] https://velvetshark.com/ai-company-logos-that-look-like-butt...
[1] https://logos-world.net/sierra-entertainment-logo/
Damn, for just a moment I thought the Sierra Online company was coming back. I want a new official Quest for Glory game.
Assume Sierra owners are too young to know what Sierra games means. I was absolutely obsessed with their logo (1) at school time.
1. https://preview.redd.it/remember-sierra-games-1979-2008-they...
https://www.youtube.com/watch?v=IMQi7olp-tw
It occurs to me just now that the logo is in fact a mountain and not a breaching Orca whale like I'd always thought
isn't it the yomesite's half dome, on their logo?
looks like it! didn't know about yosemite's half dome but that makes perfect sense considering it's in the Sierra nevada mountain range. It's gonna have to stay an orca in my headcanon though
Probably many are ashamed to remember that the name Sierra is also associated with the Leisure Suit Larry (1) games.
1. https://en.wikipedia.org/wiki/Leisure_Suit_Larry
Why would we be ashamed of that? The early Leisure Suit Larry games are a lot of fun; yeah the humor is crass and low-brow, but that's sort of the charm. It's meant to be silly.
That's because you were probably the right age to know the answers to the age control questions, and I was at an age where I could only download them from a BBS.
I actually didn't play the game until I was fifteen, in ~2006. I didn't know the answers, but I found out you can just hit ctrl-alt-x and skip the questions.
Take a look at how it was in 1995. InterAction (Sierra Online) magazine [PDF copy]. Article about the Larry 6 release on page 50.
https://sierrachest.com/gfx/Publications/IA/IA_8_1/023_Inter...
Whoa, the Vivid 3D Pro on page 62 looks awesome!
I wish I had known that! Guessing and trying the answers worked too, given no internet and only having a faint idea that "age control" was not in fact part of the game itself. I learned that Bonnie and Ronnie was not in fact a thing. What is "Bonnie & Clyde"? Eh, probably some band name was my guess. It took some patience, but since it was one of 3 games I had somehow acquired (how exactly is a lost memory), I had to get past the starting quiz.
Since I also barely spoke English at the time, I got stuck in the game itself pretty soon anyway. Didn't manage to figure out how to say some things the right way. "Ken sent me" is the last thing I remember from it... and I never had any idea that the game was rather dirty until much later.
Wow seeing that hit me surprisingly hard. Such good times
The Psygnosis logo similarly has a special place in my memory
It's simply the best looking game company logo for me.
Capstone - the pinnacle of entertainment
What was wrong with their logo? Or did you mean to type "obsessed"?
Obsessed, correct.
One of the most beautiful game logos, going back to the early nineties.
At least I'm not the only one:
* https://en.wikipedia.org/wiki/Sierra_Entertainment
I was about to post a snarky comment along the lines of "Sierra? The publishers of Homeworld and Homeworld 2?"
Every single word domain seems to have become some new AI company.
I still get briefly confused when I see a post on here about X, only to realize they're talking about Twitter, and not the display server.
There are 26 letters and millions of words; people should choose other ones.
and not only one https://www.cyera.com/press-releases/cyera-raises-400m-to-me...
Likewise, I was hoping for more Space Quest :(
Not exactly the same, but SpaceVenture finally released at the end of last year: https://store.steampowered.com/app/1374960/SpaceVenture/
This looks great but wow, what a hilariously shameless knock-off! :D
Use plunger.
That series is over, and the magical feeling of being in an open-ended fantasy world is really hard to replicate when we're not kids anymore. Loom is another game that gave me that feeling.
But there was one idea in QfG that I wish more games would use. Namely, designing three different solutions for every problem the player is facing. This idea works so well to create a sense of possibility in a game, I don't know why it got forgotten.
Me too, now I am suddenly wanting Space Quest 2026!
One can dream.
Same. A new Gabriel Knight would be fun!
Oh yeah! For some reason I was convinced that Gabriel Knight was LucasArts but nope, just misremembered.
Gabriel Knight was awesome, I'd love a new one.
Space Quest IV: Roger?
So you want to be a hero?
Well, if MicroProse could do it...
A new Lode Runner when
EDIT: holy shit I stand corrected: https://en.wikipedia.org/wiki/Lode_Runner
It’s interesting that the example interaction they use on their homepage is a no-friction example that can be handled without an AI chatbot. Why not something more complex that properly demonstrates the value?
And why do the numbers sound impressive but the words are verrrry specific (authenticating a new ... what?! What does that even f'ing mean?!)