How can AI companies ride the wave of interest without wiping out?

Tech companies have shed 151k jobs in 2023. What can AI startups learn from them?

It’s been a year of career assessment, to say the least, if you work in tech.?

According to CrunchBase, 151k jobs have been shed from the tech workforce this year—that’s not even including the bloodletting that occurred in Q4 of last year when many of the layoffs began.?

Despite ongoing layoffs in the tech sector, AI companies are raising capital like it’s 1999. The quarter ended with AI companies securing the majority of top funding deals , with InflectionAI , the year-old maker of personal assistant chatbot PI, closing $1.3B in a single round. That’s a lot of investor exuberance when the reigns are being jerked back in most other areas of venture capital.

Big Techs that made big bets in such areas as crypto, AR, and the Metaverse have laid off workers in unprecedented fashion (with some now back-pedaling with new AI-infused strategies to match up with AI frontrunners).

And yet, haven’t we seen this movie before? Companies riding an immense wave of tech transformation, kept atop the crest of consumer and investor sentiment by unforeseen forces—recessions, pandemics—before being overwhelmed and washed over?

Those of us who are still employed in tech are clutching our creaky rowboats, throwing life vests out to friends who wiped out in previously safe Google waters, or while launching their own Me coins, minting their NFT body art, building?side businesses on Zoom.

What’s someone who has made a career being a tech opportunist -- I mean, maximalist -- to do??

If you are in the fortunate position of being inside the AI bubble at this moment, looking out, I’ll bet you can hear it stretching. You life is about to get hectic; maybe it already has. In any case, take a deep breath and a moment to think about your fallen comrades in struggling emerging tech companies. And learn from them.?

Some things to bear in mind:

Get your stories straight.

As I like to say: Startup growth playbooks need a rewrite, but not entirely. Your mission from the inside is to help others on the outside see the possibilities behind AI that your application/platform/bot makes possible, and enable them to tell your story for you.?

But beware: While ChatGPT and tools like it make experimentation and discovery possible, these tools also make possible misuses a threat. And with AI, threats can have incalculable, global implications. Your job is to find, encourage, and tell stories that show how your tech made a positive difference and to surface and address the misuses before your founder is asked to testify before Congress.

What controlled, deliberate, not-stupid use cases can you seed in the market now that you can actively showcase??

I know this thinking runs counter to how many people in tech think, who prefer to “let the market tell us how to use the tech.” I find this libertarian approach an abdication of responsibility for what you are creating in the world. This bet-hedging philosophy of product development, with no thesis behind its desired impact, or emergency brake mechanism should that assumption prove faulty, is so 2018 (Two words: Cambridge Analytica ).

Speak Human.?

Speaking of 2018, back then I was leading efforts to build adoption of decentralized technologies (blockchain), and the industry was still in what I call a geek phase, seeking to gain adoption and understanding beyond a core developer crowd who could tell you what terms like “Layer 1,” “HODL” and “sharding” actually meant.?

While we had no shortage of smart people to work with who could code in Solidity, a programming language for developing smart contracts on the Ethereum blockchain, we had little to nothing to actually show the world about why Ethereum mattered beyond its then booming market price.

Ethereum was in a pre-adolescent resource-sucking stage of its development, and blockchain in general suffered from bad UX. To the average consumer, buying a Crypto-Kitty, which was then one of the few working, demonstrable if not frivolous, use cases of a blockchain application, was non-intuitive and required a working knowledge of crypto wallets and blockchain types and compatibilities. Living, successful, measurable use cases were hard to come by.

But there were a few: A program for helping unbanked citizens in the Philippines build credit through decentralized banking… A fashion line that sourced its textiles from fair trade vendors validated on the blockchain… An anonymous dating app with validated identities… It didn’t matter WHAT was being built, just that it was in the world, leveraging the tech, and improving.?

Even surfacing these attempts at applying the technology for good enabled us to build a narrative apart from the aggrandized crypto-exchanges and bought some time for the tech to catch up to its promise while staying on the radars of the consumer, municipal, and corporate sectors.

Get a good storyteller.?

In earlier stages your best evangelists are at the ground level of tech—a founder or founding product leader, who understands the tech inherently. However, startups shouldn’t leave it at that for long. While founders can make passionate and enthusiastic storytellers, they often see the possibilities but not necessarily the dents in their creations. They often see their product as the solution to everything, even when it isn’t.?

Your most reliable storyteller has to be willing to share the good, bad, the possible, and the ugly.

My foray into the world of evangelism was as a founder, nearly 20 years ago, when social media was an oddity gaining momentum. I spoke dozens, maybe hundreds, of times to corporate marketers, agencies, associations, and publishers about the space, getting in words about my product and company … eventually, if at all. More than anything I talked about what approaches worked and which ones didn’t. I made them comfortable with this new practice area that made them deeply skeptical. What they all really wanted from me was to understand this emerging space and not screw up in it. And for providing that information they trusted me, and by extension considered my company a resource.?

I believe that AI will have far wider impact—both positive and negative—than social media ever will. It’s understandable that your target customers are both curious and terrified at the prospect of integrating it before understanding it.?

Who will cultivate a bullshit-free zone to address the inevitable concerns, collaborate on best practices, educate the ecosystem, and poke holes in your messaging before regulators and press do?

That’s your storyteller.?

Overcommunicate.?

WAY back in crypto 1.0 days, I signed up for a service that boasted it would put all of my proprietary data from platforms such as LinkedIn and Facebook onto the blockchain. The website used a lot of crypto jargon and messaging that sounded right at the time—“data sovereignty,” “take back your data”—and so I clicked on a trial link, hoping to learn more. What I got were prompts to allow the system to, in essence, scrape my LinkedIn data and post it to the blockchain.?

I had expected a guided tour by a crypto-therapist-bot that would intuitively guide me through the process, explaining how my data was being encrypted and stored on a blockchain, and how I could access it moving forward. No such luck. I clicked OK to start the process and never so much as received an email confirming the transaction was complete, or what I as a new user could do to continue engaging with the tool.?

I felt like I had just been informed that I had a data doppelg?nger living in a virtual world, like in "Tron", but without the glow-in-the-dark frisbees, and yet glowing brighter than a 1,000 kilowatt LED, having just sucked up my thousands of LinkedIn contacts, yelling “Sucker!”

I felt confused, kind of stupid, and violated. Words you never want to see in your user feedback logs.

This sounds like an extreme example, but in the past two weeks I’ve checked out AI applications with interesting but unclear value propositions, asking me endless questions (including how I brush my teeth), with no context for how the information will be used or cues indicating where I am in the process. After 40 minutes of mindless multiple choice I had to call it.?

AI FOMO will only attract and keep so many users. You need to guide them, continually engage them, and process their feedback. Build communication loops immediately and use them, even if the message you are imparting is, “We’re not ready to show you anything that cool yet, but here’s what you can expect to see soon…”

Set rules of engagement.

As I write this, regulation requiring third-party audits of any corporate hiring solutions that use AI is about to take effect in New York City, and eventually elsewhere. That’s because AI, as effective as it has been for expediting massive, menial tasks such as resume filtering and critical functions such as removing human bias from the hiring process, can also be applied in counterproductive ways too. For instance, while AI can widen and diversify the talent pool of consideration for a role, it cannot be applied to the actual selection of a hire. That is an illegal use of AI.

Toe-MAY-toe, toe-MAH-toe, you may say. This affects me how?

As any martech or CRM company will tell you when GDPR regulation launched, if you are not getting in front of regulatory activities, or even leading them, you may end up afoul of them and lose much ground catching up.?

In the case of my social media startup, getting in front of regulation meant developing best practices for disclosing relationships with advertisers on social platform posts and even providing guidance to the Federal Trade Commission (FTC) on how this could be done effectively.?

Our leadership in this regard paid off. We passed through corporate compliance checks, and in some cases were asked to collaborate with organizations on their disclosure policies. It closed the loop of trust. It closed business.

Pace Yourself.?

I wrote a series designed to help startups that don’t have the luxury of massive funding to waste on growth marketing tangents, but I learned many of the lessons in that series from startups that did.?

You can’t force stardom, just ask Clubhouse , which did so much right but also rode a wave of circumstantial fumes and a16z funding without properly building a sustainable business model.?

No, all the funding in the world can’t buy you sustained market share, or product market fit. Hopefully, it buys you time to establish these and go through the required paces to stardom.

I’ll refer you here, where I outline many of the ways well-funded startups spend money before thinking . Read the whole series for thoughts on how to build optimally and sustainably.

Jory Des Jardins is a fractional executive, startup investor and advisor who has worked with B2B, B2C, and B2B2C companies on go-to-market, scale, and exit strategy. She co-founded BlogHer, a startup that achieved category-leading scale and was acquired in 2014, and is on a mission to help underrepresented founders and democratizing tech change the world – one stage at a time.