AIGC's first wave of layoffs uncovered the cracks in AI big model entrepreneurship

Original source: Wisdom

Author: Cheng Qian

Image source: Generated by Unbounded AI‌

In just one year, the AIGC industry is experiencing the ups and downs that other industries may take several years or more than ten years at an N-fold speed: from phenomenal products becoming popular, industry enthusiasm fully ignited, entrepreneurial financing emerges endlessly, to a group of star unicorns being exposed to difficulties and layoffs to survive. The speed of change is unexpected.

Myths are being busted far faster than expected. Founded 18 months ago, Jasper, an American AI chat robot startup that quickly became a unicorn and was still in full swing at the end of last year, suddenly issued an announcement in the middle of this month: The industry has undergone tremendous changes. In order to have the motivation and focus to fully seize the opportunity, it will lay off the earliest personnel in the industry to build, market and support artificial intelligence products.

This made the industry both surprised and embarrassed: Why did this star unicorn, which launched a phenomenon-level AI chat robot product earlier than OpenAI, earns tens of millions of dollars a year, and just won 100 million-level financing in October last year, choose to lay off employees when it was "not short of money"? What kind of last resort is hidden behind it?

Discussion about the reasons behind its layoffs has continued to heat up over the past week. Some people blame it on Jasper’s heavy binding relationship with OpenAI, some people think that startups have no advantage in the generative AI competition that belongs to the giant game, and some pessimistic arguments regard this as the prelude to the generative AI layoffs.

Is Jasper's layoffs an accidental case, or a typical sample that reveals the general predicament of AIGC entrepreneurship?

01. A game destined not to be won?

Why did Jasper embark on the road of layoffs? There are different opinions on the real reasons behind it.

Judging from the announcement of layoffs, the main reason is that the industry is changing too fast, technology is developing rapidly, consumer-grade AI tools are exploding, and the landscape is full of variables. The layoffs are to concentrate and adjust resources to seize the urgent emerging needs of marketing teams around the world to quickly adopt AI. It is reported that Jasper product chief Jeremy Crane (Jeremy Crane), who has served for less than a year, has left this month.

▲Jasper CEO Dave Rogenmoser (Dave Rogenmoser) announced layoffs on the company's official website on July 12

To put it bluntly, the market competition is too fierce, and it is necessary to save money and concentrate forces to capture the marketing AI market.

Over the past few days, the enthusiasm for revelations and analysis of relevant triggering factors has not diminished. In summary, the main points include the following three points:

1. Products lack differentiation advantages, making it difficult to compete with technology giants

One type of view believes that large-scale models and generative AI are, in the final analysis, giant games, and it is difficult for start-ups that lack huge capital and computing power resources to have the motivation to continue to invest in fierce competition.

Although Jasper developed commercial chat robot products earlier and built an ecological community, the competition in this field is becoming increasingly fierce. Jasper's products are easy to be copied and lack differentiation advantages. Facing the strong offensive of leading players such as Microsoft, Google, Meta, and OpenAI, they lack the ability to confront them head-on.

▲Netizen’s point of view: The degree of differentiation of Jasper products is still far from similar products.

Market insight company Synaptic believes that Jasper's products are difficult to compete with products such as Microsoft's Copilot and Adobe's image and video editing tool Firefly.

▲ Netizen’s point of view: Jasper is difficult to compete with Microsoft’s Copilot and Adobe’s image and video editing tool Firefly

2. Severe binding with OpenAI, the road to "GPT-3 shell" is difficult and long-term

Jasper's paid chatbot product for marketers is built on OpenAI GPT-3, which can also be said to rely on OpenAI to make a fortune. In the past, it was possible to occupy the market by first-mover advantage. When the stronger ChatGPT based on GPT-3.5/4 debuts, Jasper's products will only be rubbed against the ground.

Developer Sam Hogan said bluntly that Jasper is essentially OpenAI's "simple packaging", using GPT to provide services to users. Some people also vividly described Jasper's layoffs as "a short story of ChatGPT victims".

▲Developer Sam Hogan's point of view: Jasper is a "simple wrapper" for OpenAI.

The bottom line is that Jasper does not have the technical armor of self-developed basic models from the bottom, and it cannot make differentiated innovations in applications**, so its advantages established by being the first to enter the market can easily be equalized. For Jasper, a big problem is how to convince users that using Jasper service is a better choice than other ChatGPT services. Its early users have mentioned that Jasper's few neat hints and ready-made forms don't justify the subscription cost in the long run.

3. Affected by the general environment, the hype of generative AI cools down

From the perspective of the general environment, Jasper's layoffs may be an early warning of the cooling of the overall generative AI boom. With the influx of AIGC applications such as AI painting and AI chatbots with similar appearance and functions, people's sense of novelty for these applications has gradually dissipated, and more attention is paid to their actual application effects. AIGC players need to prove the unique value that their products can generate.

▲ Netizen’s point of view: Generative AI hype is cooling down.

Looking back at the time when Jasper came, its rapid rise to popularity and its dejected defeat have also sounded the alarm for many AIGC startups.

02. A year and a half to become a unicorn: Success is also OpenAI, failure is also OpenAI

Founded in 2021, Jasper is one of the fastest AI start-ups to become a unicorn. Its marketing copy generation tool makes it one of the first batch of pioneers to test the commercialization of generative AI.

These achievements are inseparable from the support of OpenAI GPT-3.

In 2020, Proof, a marketing company founded by Dave Rogenmoser (Dave Rogenmoser) and his college classmates JP Morgan and Chris Hull, with a dream of wealth and freedom, fell into crisis and was subsequently acquired by Y Combinator.

With the help of Y Combinator, Jasper obtained the priority to use GPT-3, so it launched a chatbot product based on GPT-3 in early 2021. This product can help marketers generate copy, as well as help them brainstorm product names, tweets or refine long-form article bullet points. With this product, Jasper's revenue will reach US$30 million (equivalent to approximately RMB 216 million) in 2021, and it is expected to double its revenue in 2022. At that time, OpenAI only earned tens of millions of dollars by relying on the API licensing fees of companies such as Jasper.

Before the emergence of ChatGPT, Jasper's services were almost unconditionally praised by customers, and web developers began to use Jasper instead of recruiting freelancers to write web copywriting. Some users even claim that Jasper is as good as or better than newcomers when it comes to basic writing tasks.

A large number of talents and funds have poured into this star start-up company.

The number of Jasper employees has rapidly increased from 9 in January 2022 to more than 160 within 10 months. In October 2022, Jasper received a new round of financing of US$125 million, and became a unicorn only 18 months after its establishment, with a valuation of US$1.5 billion.

Subsequently, Jasper began to focus on tailoring templates for marketers, such as generating blog posts, press releases, brainstorming product names, generating viral tweets, etc. At the same time, in order to expand the use of its products, the company also plans to make Jasper an extension of Google Chrome and so on.

However, when Logan Moser was still looking forward to letting "Jasper help users write every word", a terrible enemy suddenly arrived——

On November 30, 2022, OpenAI officially launched ChatGPT, a free AI chat robot product based on GPT-3.5.

This heralds the beginning of a new era and rings the bell for Jasper's decline.

03. Hard to beat ChatGPT's low-price offensive, Jasper's website traffic plummeted all the way

No one expected that ChatGPT would sweep users all over the world at an unprecedented speed.

According to a research report released by UBS, ChatGPT has reached hundreds of millions of monthly users in just two months after its launch, making it the fastest-growing consumer application in history.

And the more powerful the communication and generation capabilities of ChatGPT amaze users, the faster the uniqueness of Jasper products will disappear.

Chris Frantz, co-founder of a GPT-3-based content marketing startup like Jasper, once said that giving users hints is the entire value proposition of a startup like his, but now it's suddenly gone.

ChatGPT based on GPT-3.5 is available for free; even for the paid GPT-4 version of ChatGPT, the user's monthly subscription expenditure is only $20**.

The Jasper subscription fee starts at $39 per month.

▲The subscription price displayed on Jasper's official website

Within a few days after ChatGPT was released, Logan Moser quickly mobilized a team to formulate a response plan and launched Jasper's own chat function, but from the actual effect, it still failed to withstand the storm caused by ChatGPT.

As a result, users' attention to the Jasper website has been declining. In June of this year, the traffic of the Jasper website has dropped by about 50% compared to the first quarter, and it is much lower than the traffic when the Jasper service was first launched in October 2022.

Jasper's layoff plan is also on the horizon. Jasper has grown its headcount by about 140 percent since it raised $125 million last October, according to data from market insights firm Synaptic. Since March this year, the number of Jasper's new job recruitment has plummeted until June this year dropped to 0. According to its official website, Jasper currently has only two positions left.

▲ Changes in the number of newly recruited positions in Jasper from October 2022 to June this year (Source: Synaptic)

Jasper, who met ChatGPT on a narrow road, has already had a bloody encounter.

04. Conclusion: Standing on the shoulders of giants, how far can AIGC go on its entrepreneurial journey?

Can Jasper, who has been defeated like a mountain, regain his vitality and return to glory? As far as the current situation is concerned, if the previous R&D and business models are continued, most people will give a negative answer.

When the wave of generative AI first came, as one of the best AI text generation tools, this star start-up company was cheered by users and stopped by capital. And when OpenAI, which has strong technical strength, made a move, the market advantage that Jasper occupied first became so vulnerable.

The failure of Jasper reflects the hidden worries of AIGC application startups.

The basic foundation and platform service products launched by leading companies such as OpenAI and Google can become a stepping stone for innovative applications, or an accelerator for homogeneous competition. Grasping the pain points of vertical scenarios and cutting into the rigid needs of professional users sounds like a feasible idea that has been tested repeatedly in history, but how to forge a "killer tool" that is difficult for companies to replicate and become the first choice for commercial customers or consumer users, there is no generally accepted answer.

The AIGC competition has moved to the next track. It is time for Jasper startups to think about the way to survive in the future.