Jan 8, 2023

What I’m Watching in 2023

The AI, big tech, VC, and crypto trends I’m keeping an eye on this year.

Artwork by 
Anson Chan
Back to all

Brought to you by Sardine

Scams and fraud are a tax on revenue in a time where every dollar matters.  

Last year the losses to scams and fraud increased by 70%. Businesses don’t realize they have a fraud problem until they see users they thought were “verified” scamming them.

Like any skill, getting better at detecting fraud can take a decade or more of experience and requires significant training data. And it’s a problem that is constantly evolving. 

Another option? Use Sardine. 

It’s like having the best fraud team in the world, available as an API. Sardine uses machine-learning, extensive transaction data, and intelligent rules to stop fraud before it happens. Companies like Brex, MetaMask, AtoB, and Blockchain trust Sardine to reduce fraud by as much as 700%. Better yet, thanks to its superior risk modeling, Sardine also improves conversion rates, reducing the number of good customers you turn away. Thanks to Sardine,’s conversion rate improved by 10% (and it’s still going up). 

Take control of your survival and scale without the scams by implementing Sardine today.

You can listen to an audio version of The Generalist on Spotify or Apple Podcasts.

If you only have a few minutes to spare, here's what investors, operators, and founders should know about what I’m watching in 2023.

  • A big launch. Apple is expected to launch its augmented reality (AR) headset this year. It will represent the company’s most significant new product since the iPhone in 2007, and should reveal the current ceiling for AR consumer hardware.  
  • Giants show weakness. The last decade of the internet has been dominated by advertising giants like Google and Facebook. In 2023, we might see both of them challenged more aggressively. AI is opening up the search design space, while the combined frailty of social media’s giants leaves room for challengers.
  • China’s chips. Toward the end of last year, the U.S. made its most aggressive move yet to curtail China’s access to high-end semiconductors. Thus far, the response has been muted, but in 2023, we should get a sense of how Xi Jinping will seek to ensure access to this critical technology. 
  • New rules. A caveman’s summary of 2022 might be something like, “AI: good. Crypto: bad.” Indeed. If crypto wishes to avoid similar catastrophes in the future, regulation is necessary. Even with a divided Congress, the U.S. can take steps to implement better rules.

Lenin’s maxim about time and history is often repeated: “There are decades where nothing happens; and there are weeks where decades happen.” 

Though a compelling turn of phrase, our modern era seems to have rendered it partially insensible. Certainly, we imagine weeks stuffed with activity and significance – we have all very recently lived through them. But can we really visualize a decade of blankness? Ten years in which nothing of consequence occurred?

Perhaps it is simply the arrogance of the living, but it feels as if we are living in an age of hyperactivity and hypersignificance. Major scientific breakthroughs and technological advancements are announced on a daily basis. We live in an overwritten television series with too many plot points, an excess of twists and turns. 

Amidst abundant activity, it’s hard to know where to direct one’s attention. Which meaningful advancement matters most? What radical shift may have an enduring impact? 

As I consider the year ahead, there are seven areas I’m keeping a close eye on. These are topics or trends that I think have the chance to substantially alter the tech and venture capital landscape. 

Apple’s AR moment

Augmented reality (AR) has been heralded as the next great computing platform for some time. Nearly a decade on from the ill-fated introduction of Google Glass, the technology is in a strange place. 

It is ubiquitous, laminating our lives with floating flowers, hovering butterflies, striking cosmetics, and neon lettering. Social media filters have become essential to digital communication and self-expression. And yet, despite its pervasiveness, AR feels rather insignificant. Sure, we can transform ourselves into zombies and angels or fill our background with iridescent bubbles – but is that it? Beyond unconvincing e-commerce applications (Look at these shoes on your feet! Try these virtual sunglasses on!) and niche enterprise deployments, AR hasn’t approached comprehensive mainstream adoption. 

This year will help us understand how far away we are from that future. Sometime in 2023, Apple is expected to debut its much-anticipated AR headset. It’s unlikely to be a mass-market hit: there will be undoubted kinks to work out, and the expected $3,000 price point makes it a luxury purchase. While that will constrain its initial impact, Apple’s magic ski goggles will illustrate what the world’s best, most opinionated consumer hardware manufacturer considers the optimal approach. 

Already mooted design decisions are attracting attention, including Apple’s choice to rely on a waist-mounted charger connected to the device via a power cord. Seeing selections like this in the wild will be interesting – and productive. What are the limitations of Apple’s decision? Where are the obvious upgrades needed for a v2? The Cupertino company has a history of releasing compelling but imperfect new devices before turning them into must-haves. Barring a catastrophic response, the headset represents the start of another such journey for Apple. 

If the headset proves even a modest success, I expect we’ll see a sharp uptick in venture for companies in the sector and an influx of new entrepreneurs. Though AR is no stranger to a hype cycle, the presence of dedicated Apple hardware could make a meaningful difference in adoption.

Moats in AI

I don’t think we’re close to the top of this AI hype cycle. The speed of innovation – and the impending reveal of GPT-4 – will feed further frenzy. Much of this excitement will be justified, but plenty of speculation will accompany it. 

I suspect AI investing will have an especially low hit rate in the short-to-medium term. In large part, I think that’s because investors are mostly punting on a fundamental question: how do generative AI companies create defensibility?

Many of the sector’s fastest-growing and best-capitalized businesses derive the majority of their value from publicly available AI models. The release of ChatGPT has shown how tenuous that positioning is. 

Take Jasper, for example. Prior to OpenAI’s newest release, the automated copywriting platform was approaching a $60 million run rate and had pulled in $131 million from investors like Coatue and Insight Venture Partners. It had done so in less than two years of operation, with no technological edge. Much of the 160-person team reportedly focused on building new, customizable templates used to generate blog posts and Instagram captions. 

Then ChatGPT launched. Suddenly, Jasper – and its many peers – were competing with a slightly smarter, more accessible, and entirely free alternative. In response, Jasper released a chatbot interface of its own. 

Many successful companies do not have technological advantages. Indeed, as OpenAI’s Sam Altman remarked in response to a question about Jasper’s defensibility, few Silicon Valley businesses have a tech moat. They do, however, have many other forms of defensibility: distribution advantages and network effects, for example. It will be the task of Jasper and other players to find them. For now, it is representative of many players in the space: thin-product layers sitting atop powerful, external models. These are not AI companies, in the same way that DTC brands were not tech businesses.

Over the next year, I will be interested to see how tech parses this type of startup. How do VCs underwrite companies with high-potential but fragile value propositions? What other moats are most tractable? What successful playbooks will emerge for others to follow? 

Social media shakeup 

If you were forced to pick a headline story for the 2000s so far, it would center around social media. Nothing else has had such a profoundly jarring impact on the human race over the past twenty-three years. Though dominant companies have provided entertainment, connection, and information, they have also encouraged polarization, catastrophically damaged the mental health of our younger generations, and established a fractal truth environment

Doing so has been good business! Platforms like Facebook, Instagram, Twitter, Snap, and TikTok have grown rapidly and achieved massive scale. Not all have thrived in the public markets, but they are undoubtedly entrepreneurial successes. More consequentially, they have dominated our time and attention, forming an oligarchy of engrossment. 

The disruptors may be about to be disrupted. We have not seen such collective vulnerability in the better part of a decade. Meta is burning money on its legless metaverse; Snap has struggled to turn attention into revenue; TikTok is a geopolitical nightmare already banned in India and perhaps on its way out of the U.S.; and Twitter is a corporate embodiment of our tendency to see genius as boundless. 

None of these giants will disappear in 2023 and perhaps they will turn frailty into new strengths. But the rise of insurgents like BeReal, Mastodon, Poparazzi, Locket, Post, Hive, Fizz, and others suggests consumers are willing to jump ship. 

The challenge for these startups will be converting their novelty into an enduring user base. Social media companies often begin with a gimmick (think: limited character tweets or disappearing pictures), but long-term survival depends on turning that initial glimmer into an audience that keeps coming back. Another year of operation will give us a sense of just how repeatedly alluring these players really are. 

I will also be intrigued to see what business models emerge. Advertising was the lifeblood of the last generation, but there’s a feeling that it's no longer as economically or socially attractive. Apple’s privacy changes have hampered the efficiency of in-app ads, and consumers are warier of their attention being hijacked. The introduction of Twitter Blue and Snapchat Plus illustrates that even big players are thinking about subscriptions as a source of supplementary revenue, but shifting the center of gravity at either company looks unlikely. Could newcomers crack the subscription code? Or, perhaps, find another way?

Experiments in search

In the wake of ChatGPT’s release last November, it became popular to announce the demise of Google. Who needs a search engine when you can rely on an all-knowing, sonnet-spinning AI?

Though I believe reports of Google’s death are exaggerated, the search category finally has a little jeopardy. Google’s search engine has been perhaps the monopoly of the internet age – a money-spinning utility with market share that occasionally crests 90%. Despite Alphabet’s many business lines, search remains the constellation’s heart, bringing in $149 billion last year. 

Though products like ChatGPT are still limited and Google has formidable distribution advantages, new AI models open the door for competition. It will be interesting to see how, exactly, insurgents leverage these budding technological capabilities. Could we see search engines that outcompete Google in certain verticals? Chatbots that can provide quick answers with a few hotkeys? The design space suddenly looks far more open.

I’ll be interested to see how ChatGPT evolves over the course of this year. Will it remain free? Will it be able to answer questions that require knowledge of current affairs? Will its integration with Microsoft’s Bing resurrect a platform often viewed as the butt of the joke?

It will be equally fascinating to see how Google responds to a narrative that is gathering speed. Historically, the company has guarded its AI capabilities closely – will it give consumers more opportunities to witness and leverage its powers in 2023?

Battles along the silicon front

As we outlined in our piece on TSMC, semiconductors are one of humanity’s most complicated creations. They are also among the most geopolitically important. Control over their production is a strategic priority and flash point between the U.S. and China. 

Recent skirmishes along the silicon front suggest things are heating up. In October 2022, President Biden severed China’s access to leading-edge chips and the tools needed to make them. It is the U.S.’s most aggressive maneuver to date and could prompt a reaction. China may respond with sanctions of its own or take America’s rebuff as proof that it must invest more aggressively in its own semiconductor supply chain. The absurd difficulty of advanced chip manufacturing means any such investment would take years – if not decades – to bear fruit. 

One possibility is that China focuses on producing lower-end chips. By increasing supply and subsidizing manufacturers, the country could effectively undercut Western providers in this segment. In time, that may result in even American-allied businesses purchasing low-end chips from China, creating leverage.

Semiconductors are a high-stakes game with global consequences. The next few moves between China and the US may be particularly important. 

Crypto regulation

At the start of 2022, if you were asked to concoct a series of calamities for crypto, you would have been hard-pressed to have come up with something as disastrous as what happened. Disappointing, shocking, and disgraceful in equal measures, crypto showed the worst parts of its nature over the past twelve months. FTX, Terra, 3AC and other collapses revealed the sheer extent of fraud, speculation, and unethical behavior that besets the sector. I suspect we’ll see plenty of continued fallout from these debacles in 2023. 

Crypto is at a crossroads. In the parlance of Carlota Perez, we might consider it the industry’s “turning point” – the key period in which a promising technology makes necessary changes to reach mainstream adoption or dies on the vine. Though insiders will be critical to reforming the industry’s cultural norms (particularly the tendency to treat every aspect of it as a casino), external involvement is essential. Crypto badly needs regulation. There will be no “deployment phase,” no widespread adoption, without better rules. We need to crack down on fraud, keep customer funds safe, and bring more institutions into the fold. Such oversight may be anathema to crypto’s most libertarian elements, but there is a real risk of its potential being squandered without it. 

Saying we need to keep customer funds safer is one thing – finding the right legislation is altogether harder. Just as there’s a risk in doing nothing, there’s also plenty of jeopardy in doing too much. Innovation requires some latitude.

Given the global nature of crypto, international policies may prove most impactful. As president of the G-20 for the year, India is said to be prioritizing collaboration over crypto rules. We can hope that progress is made at September’s summit and its aftermath.

A divided Congress makes passing meaningful U.S. legislation unlikely. One alternative could be the creation of a “self-regulatory organization,” or SRO, overseen by the SEC and CFTC. This wouldn’t require new legislation to get off the ground and could avoid squabbles between agencies. A proposal from The Brookings Institution and Harvard Law School outlines what such an entity might achieve, along with its trade-offs. 

Crypto talent flows

In December 2021, Electric Capital released its Developer Report. Data from 500,000 code repositories suggested that crypto had roughly 19,000 monthly active developers. Though that’s considerably less than the number of software engineers at an employer like Google, it represented close to a doubling from the year prior. (I will be curious to see the latest figures when Electric releases 2022’s edition.)

As we’ve discussed, a lot has changed since then, and more pain may be on its way. Historically, bear markets have done little to curb the increase of developers flocking to crypto. Will that change in 2023? The reputational damage of the FTX fraud – a super-story that will continue to play out this year – may prompt departures and dissuade newcomers. 

It will also be interesting to see how talent flows change on a project-by-project basis. Solana became a major developer destination in the last two years, but it was also more closely tied to FTX than other communities.

For all its flaws, crypto is a hardy industry. It has weathered plenty of shocks before, and pullbacks like this one often strengthen the resolve of those that remain. I’m optimistic that a less chaotic market will allow pragmatic founders to build meaningful products in the space over the next twelve months and beyond.

The Generalist’s work is provided for informational purposes only and should not be construed as legal, business, investment, or tax advice. You should always do your own research and consult advisors on these subjects. Our work may feature entities in which Generalist Capital, LLC or the author has invested.