2024 Wrapped: the year in AI innovation
2024 was a big year for AI innovation. Lots of new technology, and technology actually getting deployed in new and valuable ways. A continued ability for tech-focused releases to confound and overcome expectations, a lot of talk about agents. But there’s still so much to do to get humans to work well with machines.
#1: Nobody can stop talking about agents (even agents)
12 months ago, when people talked about agents they usually meant people who helped you at airports or in call centers. Not any more. It’s unclear how many of these mentions from Google Trends were generated by humans, and how many by agents.
Figure: Google Trends analysis of usage of ‘AI agents’
#2: People stopped obsessing (just) about OpenAI.
2022 and 2023 were the years of obsession with OpenAI (and NVIDIA’s stock price). But this year we saw major tech companies actively deploy new frontier models and agentic software platforms - with a flourishing of innovation from Salesforce, Apple, Google, Amazon, Microsoft, Anthropic (heavily funded by Amazon), Meta, a range of new startups - and of course NVIDIA.
#3: But OpenAI also dropped a bomb on AGI.
On the last day of its ‘12 days of OpenAI’ teasers and launches, OpenAI dropped a bomb by announcing its o3 model, with some pretty astonishing evaluation results from early testing. Oh, and they casually dropped in to conversation that with o3 they are making ‘significant progress’ on achieving AGI for the first time.
#4: Chatbots are for videos and 3D worlds, too.
By the end of 2024, the launch of Sora and Veo 2 - and the early Alpha release of Showrunner - brought Generative AI into the world of video and 3D worlds. The adoption of these models will drive conversation around compute costs (and emissions).
#5: Companies started to use AI for (actually decent) ads.
2023 saw a flood of samey, boring AI-generated content in newsletters and ads - as marketers used early models like GPT 3.5 with generally pretty mediocre results. But this year, smart agencies and brands exploited the creative potential of AI. Pedigree’s campaign stood out for its sheer creativity - creating billboards with real-world dogs up for adoption in the local area, that are updated in real-time when the dog is adopted.
#6: But humans (especially female humans) are still left behind.
While there was a dizzying array of technological innovations, and it became harder every month to keep up with technological progress, there is still a long way to go in terms of the adoption and acceptance of AI within large organizations and among humans overall.
Speak to many people and they will say their #1 use for AI is ‘better emails’. So many technological innovations have not broken through to daily usage - which lags significantly among female humans compared to male humans.
