Analysts predict a cold shower for generative AI
CCS Insight just forecasted a major reality check for generative AI in 2024 as runaway computing costs and regulatory hurdles replace peak hype. Running these massive models is becoming far too expensive for anyone outside Big Tech. Why it matters: The boom is hitting reality, bringing slower enterprise rollout, search watermarks, and the first deepfake arrests.
The endless AI hype cycle might finally be heading for a harsh reality check.
According to a new forecast from analyst firm CCS Insight, generative AI is set to get a cold shower in 2024. After a year filled with sky-high promises and breathless product launches, the firm warns that reality is about to set in, driven by crushing operational costs, mounting complexity, and looming regulatory battles.
The soaring cost of compute
Building and serving generative models like ChatGPT, Google Bard, and Claude takes an unfathomable amount of computing power.
Right now, developers and tech giants alike are competing for the same scarce Nvidia GPUs to handle these massive workloads. While giants like Amazon, Google, Meta, Alibaba, and reportedly OpenAI are working on their own in-house custom silicon, those efforts will take time to pay off.
As CCS Insight chief analyst Ben Wood pointed out, the expense of simply deploying and keeping these tools alive is staggering. Tech giants have the balance sheets to absorb those losses, but for smaller companies and independent developers, the bills are already getting out of hand.
Regulators, labels, and the first AI arrests
Regulation is also moving from boardrooms to the real world, though not without serious friction.
CCS Insight predicts that while the European Union will likely lead the way with its landmark AI Act, lawmakers will struggle to keep up with the speed of technical breakthroughs. The legislation will likely need to be redrawn multiple times, meaning final rules might not land until late 2024, leaving the industry to rely on self-regulation in the interim.
Meanwhile, information integrity and safety issues are piling up. With ratings firm NewsGuard tracking dozens of entirely AI-generated news sites churned out with errors, CCS Insight expects internet search engines will soon roll out warning labels to flag synthetic articles.
On the crime front, the firm projects that 2024 will see the first high-profile police arrests for AI-powered identity fraud, as bad actors use accessible voice cloning and deepfakes to target banks, insurance providers, and individuals.
Why it matters
We think this forecast cuts straight through the noise of the past year.
For builders and enterprise buyers, the takeaway is clear: the era of throwing compute at every problem and assuming money will follow is ending. If you are building a product, relying on expensive foundation models without a clear path to unit profitability is becoming a massive liability.
For the broader industry, a slowdown in hype is actually healthy. It will shift the competitive battleground away from endless parameter counts and toward inference efficiency, pragmatic middleware, and concrete utility. Generative AI is not going anywhere, but in 2024, it will finally have to justify its price tag.
Sources (1)
- cnbc.com(opens in a new tab)Originalcnbc.com