Geekflare’s This Week in Business Tech: AI Race Heats Up
The past week has been abuzz with AI news as we approach the end of January. Donald Trump was sworn in as the 47th President of the United States and revoked the AI Executive Order signed by former president Joe Biden, Google also launched its Gemini 2.0 Flash Thinking Experimental Model, and a lot more happened in the world of AI.
So, if you have missed all the AI updates of the week, don’t worry. I am at your disposal. I have compiled all the major happenings of the week. Let’s get started!
Trump Revokes Biden’s AI Executive Order
Hours after the oath-taking ceremony, President Donald Trump repealed an AI executive order signed by former President Joe Biden. The order focused on artificial intelligence’s (AI) governance and safety. Trump’s repeal is aligned with the Republican Party’s platform, which criticizes Biden’s regulations as overly restrictive and detrimental to technological advancement.
As a result, the new administration is expected to promote an environment where tech companies can develop AI technologies without stringent government oversight. However, Trump’s approach to AI has sparked a debate among industry leaders and experts about its implications for safety and ethical considerations in AI.
Perplexity Launches Sonar API With Real-Time AI Search
Perplexity launched its real-time AI-powered search tool API called Sonar. It is the first of its kind, as the majority of the generative AI features today are backed by training data and don’t have access to the Internet, which limits their possibilities. To address this issue, Perplexity has launched this API, which gives a real-time connection to the Intenet and gives informed advice backed by trusted sources.
It is available in two versions: the Base and Sonar Pro. As expected, the Base version offers an economical solution for integrating AI search into applications, costing $5 for every 1,000 searches, additional fees of $1 for every 750,000 input words, and $1 for the same amount of output words. The Pro variant is tailored for more complex queries and is also priced at $5 per 1,000 searches, but it charges $3 for every 750,000 input words and $15 for the same amount of output words. Sonar Pro provides enhanced processing capabilities and doubles the citations.
Google Launches Gemini 2.0 Flash Thinking Experimental Model
Google launched its updated Gemini 2.0 Flash Thinking Experimental model under the experimental tag “Exp-01-21” via Google AI Studio and API. It’s based on the Gemini 2.0 Flash framework launched in December last year. The model has also achieved the top position on benchmarking charts by scoring 73.3% on the AIME2024 (mathematics) and 74.2% on the GPQA Diamond (science).
However, the spotlight here is that Google has kept the 2.0 Flash Thinking Experimental for free during its beta phase, which gives it a competitive advantage over OpenAI’s o1 Pro, which charges $200 per month.
OpenAI’s Next-Gen Reasoning Model’ o3 Mini’ Is Coming Soon
OpenAI isn’t far behind in the race to innovate reasoning AI models. CEO Sam Altman has announced that their next AI model, o3 Mini, will launch in the coming weeks, succeeding the o1 model introduced in September 2024.
The o3 Mini is expected to compete with Google’s latest Gemini 2.0 Flash Thinking Model. However, the launch of the o3 Mini has been overshadowed by the controversy about the lack of transparency regarding the funding of FrontierMath, a benchmark used to test the model’s mathematical skills. Critics argue that this lack of transparency undermines the credibility of the benchmark. However, Epoch AI, the organization that developed FrontierMath, acknowledges the need for greater transparency.
That’s a wrap for this week’s highlights. If you want to stay updated on Business Tech or AI, be sure to follow us on Google News. We’ll keep bringing you the latest happenings in the business tech world!