Generative AI will become widespread in 2024, taking root in laptops, smartphones, and everyday technology. The rise of multimodal models has pushed generative AI to break new ground, processing text, video, images, audio, and even delivering output that seamlessly combines these formats.
As 2024 approaches, TechRepublic revisits the year’s biggest generative AI stories.
1. NVIDIA AI architecture is sold out
NVIDIA was the clear winner in AI this year. Introduced in March, Blackwell chips have become the gold standard for GPU microarchitectures for processing large amounts of information. Blackwell enables AI training, research, and computing for Amazon Web Services, Google, Meta, Microsoft, OpenAI, xAI, and more. These processors were so popular among businesses that as of October, Blackwell chips were sold out for the rest of the year.
In March, sales of NVIDIA’s Hopper chips pushed the company’s market cap to $2 trillion. NVIDIA has become one of the three most valuable companies in the world, along with Microsoft and Apple. AMD and Intel also offer AI accelerator chips, but their businesses haven’t experienced the same explosive growth as NVIDIA.
In the broader generative AI industry, the success of these powerful processors is just one of the many ramifications of the growing demand for large-scale, high-density data centers.
2. OpenAI has shown secrecy
This year, many leading AI companies have developed enterprise products, more powerful models, and experimented with new hardware. When it comes to OpenAI, there have been whispers of a “strawberry” model that takes a further leap towards human-like intelligence. Strawberry turned out to be OpenAI o1. It is a “reasoning” model aimed at solving more time-consuming and trickier problems than its predecessors in the GPT-4 family.
3. AI PCs become mainstream
Technology enthusiasts may look back to 2024, when AI became a standard feature on nearly every new PC. From Apple Intelligence to Microsoft Copilot, embedded AI has been everywhere. Gartner predicted in September 2024 that by 2025, 43% of all PC shipments will be AI PCs.
However, a November article in Reuters found that while backorders for NVIDIA’s AI processors may be limiting the availability of AI PCs, overall demand for PCs remains low. I did.
4. Has AI brought new ways of thinking about UI?
Since the debut of the Apple Store in 2008, apps have been the primary way consumers interact with their smart devices. Voice assistants like Siri add an extra layer of convenience, allowing users to control certain applications through voice commands. The next step for the AI industry is to seamlessly control all aspects of PCs and mobile phones through AI, as demonstrated by Microsoft Copilot and Claude 3.5 Sonnet’s Computer Use capabilities.
Claude’s use of computers allows AI to translate natural language instructions into executable commands, such as moving a cursor, typing, and interacting with computers the same way humans would. However, this feature can be resource intensive. For example, performing a simple task like opening a URL and extracting information from a website can cost as much as $0.31 in tokens.
5. Microsoft recalls have been repeatedly postponed
Microsoft Recall has been controversial and raised security concerns since its inception due to its unrestricted access between Copilot and the rest of your PC. A public preview was scheduled for June, but early access to Recall was delayed until October and then December. Recall was intended to be the basis of the Microsoft AI PC, but Microsoft’s delay in pursuing a “safe and trusted experience” meant that Recall pushed the boundaries of what consumers wanted to share with AI. It shows.
More must-read AI articles
6. “Search” option added to ChatGPT
In October, OpenAI extended ChatGPT with Google-like search engine functionality. ChatGPT searches provide generative answers that include links to external sites. Adding up-to-date information such as weather forecasts could make OpenAI more competitive than Google search. OpenAI has entered into agreements with several media outlets to license content that appears in ChatGPT search.
Initially, only ChatGPT Plus and Team users had access to ChatGPT search. However, in December the tech giant made it available to all users.
7. Getting started with Apple Intelligence
Apple remained silent on much of the AI race, waiting until 2024 to announce its plans. Apple Intelligence was announced in June for new devices. Apple has worked with OpenAI to add many standard AI features such as summarization and rewriting. It will also restrict the creation of images, mostly in a cartoonish style, to avoid the possibility of users creating deepfakes.
Apple Intelligence uses the tech giant’s M-series chips or A17 Pro chips or newer. Starting with iOS 18.2, Apple devices can connect to ChatGPT to handle more complex questions asked to Apple Intelligence or Siri.
8. Google investigated Gemini use cases for the enterprise
Google Gemini was not a new development for 2024. However, a December 2023 launch and the release of the smaller model Gemma in February means that most of Gemini’s public happenings have happened this year. Google replaced the Bard brand with a more powerful model and introduced Gemini in search, mobile apps, Chromebooks, and the Vertex AI cloud platform. Following the “agent” AI trend, Google released custom “gems” in August.
9. AI regulation continues to evolve
In 2024, governments worked to regulate the use of AI. In Europe, the European Union’s AI law came into force in August, outlining prohibited uses and seeking to provide guidelines to encourage innovation.
The UK has created an AI assurance market to pursue businesses in the field of generative AI, and more legislation is expected to be enacted next year. The UK has joined several international initiatives with the US and others to standardize AI safety.
In the US:
See also: Seven new AI R&D facilities to open across Europe from 2026 thanks to High Performance Computing Joint Venture.
10. Mature video generation technology
Generated AI videos are still imperfect, often resulting in inconsistent scenes and creepy results like strange proportions or distorted limbs. But that hasn’t stopped companies from releasing AI video generators.
OpenAI’s Sora was first demonstrated in February and released to ChatGPT Plus and Pro users in December. Google’s Veo is available to some Google Cloud customers. Canva also offers AI video generation.
