2022 will be known as the year when a lot of AI technology became accessible to hackers and developers. With the recent explosion of capabilities as shown by Chat GPT and Stable Diffusion, it is clear that AI applications to our daily lives are going to explode in 2023. This could be one of the most exciting technological advances since the iPhone.

What’s harder to do is to make sense of the companies best positioned to take advantage of this breaking dam, what the market looks like and where it is likely headed in 2023. With core expertise in applying AI to search, using existing AI technologies, and training our own models, here is our analysis of the industry landscape and our predictions for what is to come.

Let’s start off by categorizing the AI players into a 5 classes.

These include Anthropic, OpenAI, Cohere, and AI21. All of these players will all have instruction following models that are more or less substitutes for each other. Their businesses will all be based on an API model. Their APIs will all converge to (a) generations (b) embeddings (c) fine tuning (d) classification. They will all attempt to recruit AI frontend startups (see below). Inspired by the likes of Stripe, they will use an “index YCombinator” business model, and outsource GTM to AI frontend startups. The moat for these companies will be the quality of their model,  their ability to recruit loyal customers who use their APIs.




These are companies like Jasper and Copy.AI that are vertical specific “prompt engineering layers” on top of the foundation model players. To a first cut, all these companies will look like a designer, 2 full stack engineers, 1 BE engineer, 1 applied ML engineer and a GTM team. These companies will use APIs from the foundation model players to stitch together end user solutions. Their strength will be user focus, and a relentless focus on GTM. Most of YCombinator’s 2023 vintage will be companies of this form. A large number of these companies will be marketing tech companies, and the market will be flooded with different styles of textual and visual marketing artifacts because the cost of generation of tailored messages is phenomenally lower than it used to be. These companies will look a lot like a dressed-up version of the enterprise SaaS companies from the last decade (PLG, consumer grade UX), with an “AI intelligence layer”.




These will include the original CoPilot for programmers (GitHub/MSFT), “LLM for Chatbots” (Character.ai, Inflection), “LLM for RPA” (Adept), “LLM for search” (Neeva) and many others being incubated as we speak. These will also include OG LLM companies like Cresta (enterprise chatbots) and Lilt (translation). All these companies will make a bet that task-specific LLM tech + deep understanding of user + product is how the market is won. They will need big capital raises to pull off their mission.



 These are classic shovel providers in a gold-rush and will include the labeling companies (Scale.AI, SurgeHQ, Snorkel), training infra companies (MosaicML, Stronger Compute), inference infra companies (Goose.AI), and many others.



Big compute clouds like GCP, AWS, Azure and Oracle


These companies will all wake up to the business opportunity that LLMs provide and compete very aggressively to go up the value chain beyond being just “GPU compute providers”. Azure threatens to have run away with this game already.

Over the next year, we expect to see multiple battles, consolidation, and splintering across these categories. Here are our top predictions of what 2023 looks like:


Compute clouds and foundation model players will collide. The compute clouds will all move towards making “LLM AI capabilities” a feature of their platforms. This will cause the first and last categories of companies to converge. Microsoft already has its OpenAI investment. Google will join the rungs of the foundation model providers by exposing an LLM API via GCP. Amazon and Oracle will buy out one of the independent foundation model companies, or roll out their own models. It’s an open question if there will be any Snowflake and Databricks-style independent companies at the end of 2023.



The CoPilot for X”companies will find themselves competing with AI frontend companies that use foundation model APIs to beat them to market. We’ll see this play out in search, RPA, programming and other categories. The jury will be still out on whether you need to invest in deep tech to succeed in these verticals, or can build a frontend business on top of the pure foundation API players. We think it’s the former.



The tooling companies will find the going a bit hard. The largest of the foundation models will turn out to be great for producing labeled data, often better than humans. The AI labeling companies will find themselves competing with the large models they helped bootstrap. The infra companies will find the open source tooling continues to get better and the GPU compute availability situation becomes easier, eroding their advantage. Both these classes of companies will react by “moving up the value chain” to solve for business use cases and find themselves in competition with the “AI frontend companies” and the “CoPilot for X” companies.



OpenAI will try to be both an API provider and a “CoPilot for X” company for many use cases. This will lead to channel conflict and friction with its partner companies. We will start to see beginnings of rent seeking behavior.




And finally, traditional enterprise companies like SalesForce and ServiceNow and UIPath will find themselves at the end of an onslaught the likes of which they haven’t seen before. They will be threatened by one of the “Copilot for X” or “AI frontend” companies who will promise automation to disrupt their existing business model. They will react by acquiring “CoPilot for X” companies.

With core expertise in applying AI to search, using existing AI technologies, and training our own models, here is our analysis of the industry landscape and our 2023 predictions for what is to come.

2023 AI Players and Predictions

We are delighted to welcome Steven Shure to the Neeva team. Since Neeva became publicly available in June 2021, over half a million people have visited the site. We’ve been thrilled at the response and the number of people who have signed up for Neeva. Steven’s arrival marks a critical step in the Neeva journey, and a commitment to building the business of Neeva at the same standard at which we’ve been building Neeva search itself since day one.  As the architect of Amazon Prime’s early and extraordinary growth and product evolution, and with three decades of experience overseeing widely-loved, consumer-paid subscription businesses, Steven is uniquely qualified to help the business of Neeva realize its full potential.

In the past months, we have made material investment in building out Neeva’s own web search stack to provide users the best possible search experience, private, and free of ads. Our stack is built on top of tens of billions of web pages we have selected from the hundreds of billions of URLs that our crawlers have discovered. With Steven’s arrival, the coming months will see similar investment and progress in how we take Neeva to market as well.

From the start, at Neeva we have believed in excellence. From the Neeva board to brilliant and broadly respected technologists, like Udi Manber and Darin Fisher, we have entrusted Neeva to 

 who both share a deep passion for Neeva’s mission and themselves have a history of building extraordinary products and businesses. We are thrilled to add Steven, a leader with a similar track record in subscription-based products and businesses to the team. At Neeva, Steven will serve as Chief Business Officer, overseeing all business operations.

Steven joined Amazon in 2006 as the first Vice President of the just-launched Amazon Prime, overseeing all aspects of that business including product, engineering, and marketing. He helped build the service from the ground up, including its launch across Europe and in Japan and the incorporation of key benefits like streaming video. In Steven’s seven years overseeing Prime, its membership grew from fewer than 100,000 to more than 10 million. Steven added Amazon’s Worldwide Consumer Marketing organization to his responsibilities in 2008, overseeing all performance marketing for the company for 11 years, and administering a budget of over $2B annually. In addition to Prime and Marketing, Steven held senior leadership positions in Supply Chain Technology and Amazon Web Services, managing teams from as small as ten to greater than ten thousand. Prior to Amazon, he spent two decades at Time Inc. overseeing direct-to-consumer marketing and subscription services for magazine brands such as People, Sports Illustrated and Entertainment Weekly.

Steven’s ex-colleague, Christopher North, Managing Partner, L Catterton, and former Managing Director, Amazon UK says “During the decade we worked closely together at Amazon, Steve built an outstanding team, collectively driving large-scale innovation across global marketing and Amazon Prime that helped power growth from $11B to $136B in 2006-2016 alone. His thought leadership, domain expertise, and humble style made him one of the great all-time Amazonians.”

Says Steven, “I am beyond excited to join the Neeva team. Since my start in the magazine industry, I’ve always had a deep belief in the power of customer-supported, subscription-based products, and have seen firsthand how those products (and customers) are inevitably compromised when they become dependent on advertising revenue. Above all, it is my belief in Neeva’s mission, and the extraordinary team already in place, that led me to want to join. I remember people asking me years ago why I thought anyone would ever want to subscribe to a shipping service. I hope and expect that in the not too distant future, the question ‘why would anyone want to pay for a search engine?’ will, in hindsight, sound just as ridiculous.”

We’re continuing on our journey of building the world’s first ad-free, private subscription search engine that puts the customer first, and we’re thrilled to have Steven help make that a reality. We’re excited about what’s ahead and if you share a similar passion come 

A longtime Amazon VP, Steven played a pivotal role in building and running Amazon Prime and oversaw Amazon’s Worldwide Consumer Marketing organization for over a decade.


We are delighted to welcome Steven Shure to the Neeva team. Steven played a pivotal role in building and running Amazon Prime and oversaw Amazon’s Worldwide Consumer Marketing organization for over a decade. At Neeva, Steven will serve as Chief Business Officer, overseeing all business operations.

Neeva Continues Growth, Welcomes Steven Shure as Chief Business Officer 

So much of the infrastructure that we all rely on has not kept pace with the industry’s respect for user privacy. Let’s take email campaigns as an example. When we opened up our waitlist, we needed a way to track sign-ups and communicate with our prospective users. We picked a popular off-the-shelf email provider, one with rich templating tools and solid list management. We didn’t think we would run into issues with tracking. With no third-party advertising, there should be no incentive for invasive data collection. We should just be able to just turn it off, and we’d be good to go, right?

Wrong! Despite disabling every tracking option we could find, our provider 

 insisted on embedding tracking pixels in every email we sent. Sure, seeing the open rate on our campaigns is nice, but the tradeoff was not worth it. This is especially true because tracking pixels can cause much bigger privacy issues, potentially exposing user location, activity, and even personally-identifying information if 

 are used. We just wanted to welcome our users to Neeva; instead, we wound up inadvertently intruding on their personal space. (Of course, we changed our email provider right away!)

The problem was that our original provider couldn’t imagine that we would 

 want to collect this data. Open rate is one of the fundamental metrics of a campaign’s success, and the idea that someone would be willing to forgo this information didn’t occur to them. For us, though, this tradeoff was essential: our users expect better from us.

Even more concerning, some B2B service providers reserve the right to misuse our customers’ data. We evaluated one communication provider whose privacy policy said that they were free to combine information that they collected about an email on their site with the information in 

 to those same emails. Similarly, when using URL shortener services, we could not reliably confirm from a leading provider that information resulting from Neeva users clicking and following shortened URLs would not be used for advertising! Needless to say, our commitment to user privacy must extend to every service we use. These services lost a potential client because they were unable or unwilling to commit to respecting user trust.

We don’t say this to call out particular providers, but instead to call out a 

. Every platform advertises how it will help you 

 your audience, from simple things like open rate or click-through rate to sophisticated options like audience demographics that rely on user tracking joining with third-party data. What very few of these products advertise is how they will help you 

 your audience. User data is very valuable, and refraining from questionable and invasive data collection can seem like leaving money on the table. Service providers need to work to understand the mindset of companies who see user data as a sacred trust, and not a resource to be exploited. There are quite a few of us, and as people seek services that respect their privacy, the number will only grow.

As a company, we have to put our money where our mouth is. We have been careful to pick tools that serve our needs and allow us to protect our user’s privacy, such as 

 for analytics. In cases where we have had difficulty finding options that align with our principles, we’ve built our own, as we did for collecting customer feedback and creating shortened URLs.

Our plea to the industry is to make this easier. Creating 

 services is becoming the norm for consumer services. Make it easier for us to find and buy solutions that take privacy seriously. Put data collection options up-front, and clearly explain what is being collected, where it’s from, and most importantly how to control it. As user-facing companies become more privacy-sensitive, there’s an increasing need for B2B services that share that goal.

Respecting user privacy isn’t something that can be done in the last mile. It requires us to respect our users from end-to-end and at every point we touch, whether it’s through email, social media, customer support, websites, or our apps. A chain is only as strong as its weakest link; we can’t build privacy-first products without service providers that do the same. As people become more aware of 

 and used, they will hold us to an increasingly high standard. Respecting user privacy is something all of us should embrace.

When we started Neeva, we became part of a movement of consumer-centric companies focused on privacy. We expected several challenges: how do you build a compelling service users are willing to pay for? How do you succeed in a crowded market? What we did not expect was how difficult it would be to 

We can’t build privacy-first products without service providers that do the same. Make it easier for us to find and buy solutions that take privacy seriously. Respecting user privacy is something all of us should embrace.

The need for privacy-first service providers

Yet with every passing month, there is increasing awareness and escalating concern around data privacy, 

in the core services we use in our daily lives.

That’s why we’re focused on bringing a new way of thinking to the search experience by focusing on you: the customer. Since launching Neeva, our mission has been greeted with enthusiasm and encouragement. Today, we are excited to announce three technology pioneers are joining our team and that we’ve raised an additional $40 million Series B funding round to accelerate our work.

First, I am delighted to welcome search guru 

 to Neeva. I (and many others at Neeva) have worked closely with Udi, Margo and Darin for many years, and are delighted to work with them again toward delivering on Neeva's mission.

Udi Manber is one of the most influential computer scientists behind the evolution of search. He has solved some of the deepest technical challenges in search over the last 20 years, and will be instrumental in Neeva’s work on advancing our own search capabilities.

We’re also delighted to welcome Margo Georgiadis, who most recently was President and CEO of Ancestry to Neeva’s Board. Margo has had a multi-decade career building and scaling innovative global consumer services as CEO of Mattel, President of Americas at Google, EVP of Card Products at Discover, and a Partner at McKinsey. An invaluable addition to our esteemed group of board members, including Asheem Chandna, Greylock, Reid Hoffman, Greylock, and Bill Coughran, Sequoia Capital, Margo’s perspective and counsel will be instrumental as we grow and ready our search engine for the market. I am also excited to welcome 

 from Inovia as an observer to the Neeva board.

Lastly, I am excited to announce that Darin Fisher will be joining our team. Darin joins us from the Google Chrome team, where he spent over 15 years building Chrome into the product it is today. Darin will help us redefine the interaction between the browser and search engine to better serve the end customer - our users.

Second, I’m happy to share that we have raised a new $40 million funding round, led by our original investors, 

 participating. Greylock and Sequoia have been strong advocates of ours from the beginning, driven by their belief in companies that are focused on building incredibly useful services that users can trust. Patrick Pichette’s experience at Google as its CFO, and his prior experience as Bell Canada’s CFO, make him a valuable addition to the Neeva family.

With this funding, Neeva will be able to invest more in the foundational infrastructure needed to make search much better in commercial verticals, which are often crowded with advertisements. We will also be expanding into non-US markets in the quarters to come.

We’re continuing to build a search engine that puts you first, and are making great strides towards that vision every day. We’re excited for what’s ahead and for you to experience the magic of ad-free, 

 for yourself. Please sign up to be on our waitlist at 

Search engines have done more than anything else to expand the internet and make online research, commerce, and connections possible. Much of my career has been spent expanding the usefulness of search.

The Neeva team is excited to announce that three technology pioneers are joining our team and that we've raised an additional $40 million Series B funding round to accelerate our work. With this funding, we can continue to make great strides towards building an ad-free, private search engine focused on you -- the customer.

Neeva Makes Progress in Reimagining Search

Simply put, it allowed for digital platforms to host user content without being responsible for what the content said. While there has been clarification and litigation, the essence has stood the test of time.

It is important to understand that this law came at a time when there was no Facebook, Twitter or YouTube. And there was real fear that a torrent of lawsuits would squelch the fledgling Internet.

There are a few platforms whose ability to shape conversation has become incredibly important: Facebook, Twitter, YouTube and WhatsApp have become the public squares of today. And the software that controls them has enormous power over humanity.

At many levels, these platforms are struggling. They struggle to protect democracy and civil discourse. They can barely keep out what is outright illegal. And worse, they are getting more entrenched over time, not less. Free markets are failing us, platitudes about "competition being a click away" notwithstanding.

At one level, a company is a completely unsuitable construct to police the speech of the world. Given that we live in a world of supervoting shares, do we really want Mark Zuckerberg's grandchild to decide "truth" for us? (Thanks to his supervoting shares, this is more of a predictable outcome than you think. And no, to the best of my knowledge, he doesn't have a grandchild. That's my point.)

And if you are an executive at the company, it is not clear what you are expected to do. Back when I was running Google ads, I once asked Senator Ron Wyden--who had been the lead on Section 230-- what he wanted us to do about Russian interference in the elections. He waved a finger at me and said, "Stop all the bad actors."

Most politicians, citizens, and even tech workers would agree with the sentiment. The problem of course is in how you do that.

While most people are aware that Section 230 protects online platforms from being sued about content that they host, it also has a "good samaritan" provision that allows these platforms to screen or remove content that they think is "offensive".

 making decisions in secret about issues that might literally change the world?

"Repeal 230" is a popular thing to say these days. What happens after that? A spate of lawsuits against social media companies about content that they are hosting? Does this also not mean that no new startup will come up in user-generated content for fear of lawsuits?

We need to understand the nuance of the underlying problem, and move towards some standards and shared definitions. Here are some concepts that are difficult for standalone tech companies to decide and rule on:

Should content from influential people (with a lot of followers, let's say) be treated differently than content from people with very few followers?

Is there a commonly accepted definition for hate speech or falsehood or misinformation?

How much exposure should recommendation algorithms (which are under the control of the platforms) generate for problematic content before it is subject to more review?

Now that we've laid out some topics that require more thought, what are good first steps that we should be taking to drive change?

We need common definitions. These definitions need to be arrived at by a group of people who are clearly not controlled by any individual company.

We need the tech companies to fund the creation of a shared system that can be used by all companies to generate labels for these definitions. Such a system should be available at a low cost for startups.

This shared system will make it easier for new companies to create innovative products and compete with the giants.

Whether a site allows porn or hate speech is up to them. They should clearly show what they do so that people can make choices about what product they want to use and what content they want to see.

The degree of liability that a company incurs should then be dependent on the kinds of policies that a company adopts: If a site thinks that it is OK to run unsubstantiated conspiracy theories, perhaps they should be subject to lawsuits about those posts.

We need the tech giants to create a system that can be used by all of them: Individual companies can decide what they want to allow and what labels to put on content, but it is critical that we think of this as a system that should be subsidized by the big tech companies while being accessible to everybody (startups, for example) for a low cost.

Stopping the spread of misinformation and preventing too much decision-making power from concentrating in the hands of a few companies are incredibly complex problems that can literally shape electoral as well as life and death outcomes. For us to find better answers to these problems, we need to start asking better questions. While it is clear that the blanket protections of Section 230 don't make sense for today, as with another topic, "repeal" is easy and tempting. What we need to do is work on the "replace" part now!

There has been a lot of talk about repealing 

 With some merit, Section 230(c)(1) has been called "the twenty-six words that created the Internet." (This law was passed in 1996; talk about prescience!) 

There has been a lot of talk about repealing Section 230. With some merit, Section 230(c)(1) has been called "the twenty-six words that created the Internet." (This law was passed in 1996; talk about prescience!) Simply put, it allowed for digital platforms to host user content without being responsible for what the content said. While there has been clarification and litigation, the essence has stood the test of time.

Why repealing Section 230 is not enough

One of the important questions that people asked us over email and social media was around the interplay between 

 and a personal search engine. At Neeva, we passionately believe that these two concepts are not in conflict. That means showing you weather for the locations you care about, news about the sports teams that you like and stocks that you own and want to see updates about. All while having these preferences seamlessly work on every device you own that you connect to Neeva. Bringing together personal and private creates a strong product that drives immense value for people!

This is no different than getting personalized movie recommendations to watch or playlists to listen to. 

A search engine that knows nothing about you never gets better for you!

We firmly believe that a search engine that knows you can deliver a great experience while protecting your privacy. Not only do we promise that you will never be shown ads on Neeva, we guarantee that your data will never be sold in any form whatsoever (aggregated or not) to anybody. You are firmly in charge: we delete your search history by default after 90 days. The 90 day retention period is the de-facto behavior of Neeva: based on your feedback, we have made it explicitly part of our 

. We have also added more details to our privacy policy about how we handle your data, and we plan to make further improvements in clarity and readability over the coming weeks.

In addition to a short data retention policy, we understand that you may have searches that you want to be completely private with nothing stored. We offer an incognito mode for these cases. We want to give our users options for how they search the internet for content that matters to them.

While we currently use data from other providers in addition to our own results (like using Xignite to show you stock prices or weather.com for weather), we hold ourselves accountable for the quality of the product. We are passionate product people and will build whatever we need to create a great search experience for you. We respect your privacy regardless of the source of your search results. We take an extraordinary amount of care to ensure that we do not pass on personally identifiable information such as your IP address or precise location to external providers. We are responsible for the quality of the results we show you, and we will do whatever it takes to provide an amazing product for you while respecting your online privacy.

Lastly, we have heard questions about what to expect in terms of a price point for a Neeva subscription. I talked about this a bit with Kara Swisher in her recent episode of 

. Neeva serves only one constituent - the user. This is made possible with a subscription business model. We will start at a reasonable price, say the cost of two black coffees a month. We hope that over time we will be able to reduce the price as we become more efficient. Everything that we take for granted (our computers and phones, even ice cream) started out at a higher price point before scale and innovation made them available to everyone. We hope Neeva follows a similar trajectory.

We are well aware of the responsibility we have in protecting your data, which is why we plan to publish a white paper on our security and data handling practices in the next few weeks. If you have any questions or thoughts, 

We were overwhelmed and humbled by the response to the announcement of Neeva. Tens of thousands of users signed up to join our waitlist and wrote in so many reasons why they were excited to use Neeva. Thank you all for the feedback!

What is the future of search?

I was fortunate to be part of the early engineering team that built Google Search, which democratized access to knowledge for millions of people around the world. It was, and remains, a vitally important service, made free for everyone through online advertising.

As the leader of Google’s advertising products until 2018, I truly believed in the benefits of an ad-supported search experience. And indeed, there are aspects of online advertising that I still very much believe can bring benefits to users and advertisers alike. However, I have come to believe that 

 detract from a good search experience and have also had many unintended side-effects that have large social consequences.

First, the very existence of an online ad right on top of the search results pushes down and deprioritizes the information we are searching for. After all, when most of us are searching about flu symptoms, we want to know the symptoms—not see an ad for cough syrup, and certainly not an ad for cough syrup that will then 

 across the Internet for the coming weeks.

Second, ad-supported search engines face the daily pressure of returning value to their shareholders by prioritizing advertisers and ads revenue. This has several unintended consequences, including ever-increasing ad-load, ads driven misinformation and harmful content, and practices that value profit over user privacy.

These problems are further exacerbated by the fact that there tends to be only one or two viable search engines in most countries. A lack of competition in the search space is bad for innovation and bad for users. We need more search engines, offering different kinds of experiences and preferences, to fit people’s individual needs.

Enter Neeva, our vision for a private and ads-free search engine that truly puts you first. Whether you’re looking for general information, or something deeply personal—like an important email, a calendar invite, or a copy of your passport—you can be certain that your information is your own, and not sold on to advertisers as a means of targeting you with their ads.

Neeva isn’t just about privacy, it’s also about simplicity. Most of us are overwhelmed by the vast amounts of information competing for our attention 24/7. Neeva lets you easily find just what you want, either on the public web or in personal accounts you choose to connect.

As an example of how you can use Neeva, if you’re upgrading your study (since we’re all spending a lot of time at home these days), we can help you find the best standing desk and home monitor. Neeva prioritizes quality review sites and brands you trust instead of flooding your screen with paid results. We can also help you save and organize your research so you can come back to it easily.

Neeva’s web search for “best standing desk” and saved research on topics of interest

Neeva’s Home page and search results from personal services

Search is the gateway to the world’s information, and with Neeva, we want to help you experience the Internet in a new way—free of distractions, prying eyes and frustration. Moreover, we strongly believe that you, the customer, have inalienable rights online: we published our manifesto in a consumer 

We are currently in limited beta testing. Please 

 to get early access. We’re still in the early stages of building our company and our product, and we’re excited to build it with you.

Finding information that is important to us—weather, jobs, sports, that strange headache symptom—is a basic and deeply personal human need. In less than twenty years, we’ve gone from looking up information in encyclopedias to being able to find almost anything with a tap on our phone!