Sign up for our daily and weekly newsletters for the latest in industry-leading AI news and exclusive content.
A year ago, I wrote an AI Beat column about five AI stories I was eager to cover in 2023. The rapid flood of AI developments, especially following the release of ChatGPT by OpenAI in November 2022, meant I didn’t have to wait long. Subjects like GPT-4, the EU AI Act, AI search battles, open vs. closed AI, and the demand for training data and computing power kept me busy throughout the year.
As we approach a new year, here are the AI narratives I predict will dominate in 2024:
1. OpenAI vs. Anthropic
In early 2023, OpenAI was thriving on the success of ChatGPT. However, Anthropic, founded in 2021 by former senior members of OpenAI, quickly became a significant competitor. Anthropic introduced its ChatGPT rival, Claude, on March 14, 2023, the same day OpenAI launched GPT-4. Both companies are now hunting for major funding – Anthropic around $750 million from Menlo Ventures and OpenAI at a staggering $100 billion valuation. This intense rivalry is set to escalate in 2024.
2. Open Source AI Closing in on GPT-4
French startup Mistral AI surprised everyone in early December with an open-source LLM named Mixtral8x7B, shared via a torrent link. This move signaled that the open-source AI community is fiercely determined to match proprietary models. Mistral’s CEO announced plans to release an open-source model equivalent to GPT-4 in 2024. Meanwhile, Meta is rumored to be preparing Llama 3 for release in the first half of 2024, a model expected to rival GPT-4. Meta is keenly gathering feedback from its developer community to shape this forthcoming version.
3. AI’s Influence on the 2024 Elections
Concerns are mounting over AI’s impact on the 2024 Presidential election. The interim dean at the University of Chicago Harris School of Public Policy likened it to how social media influenced the 2016 and 2020 elections. Generative AI, through chatbots and deepfakes, is expected to create significant complications. Political dynamics are likely to affect AI regulation, as the election becomes a battleground for candidate positions on AI, potential misuse, and media handling.
4. Training Data: The Core of AI
Training data for AI models has been a hot topic in 2023 and is expected to grow even more important in 2024. Issues like copyright, bias, deepfakes, disinformation, labor, and the open vs. closed model debate all revolve around the data used to train AI. Recent developments include LAION’s dataset removal and OpenAI’s negotiations to create new datasets. This area holds many critical discussions that are only beginning to unfold.
5. Effective Altruism vs. Effective Accelerationism
The future of AI is being shaped by evolving belief systems around its risks and opportunities. Effective altruism (EA) has built a strong influence in AI policy, particularly in AI security, while effective accelerationism (e/acc), championed by VC Marc Andreessen, presents a contrasting perspective. Covering these belief systems is crucial, as their billionaire backers will continue to significantly impact AI policy and investment.
Stay informed with the latest news delivered to your inbox daily by subscribing to our updates.