Observed Signal · Jan 28, 2025 · Technical Release · Source: OnlineMarketing.de · Impact: 2/5 · Sentiment: Neutral
Perplexity Adds DeepSeek R1 Reasoning Model
The article highlights DeepSeek's R1 AI model as a hot topic, with Perplexity integrating it into its AI Answering Machine and the Perplexity Assistant. R1 powers Perplexity's Reasoning Mode, enabling multi-step reasoning and enhanced search across apps. Perplexity also introduced API Sonar to embed AI-powered search into apps. DeepSeek emphasizes RL-based training and high efficiency with minimal labeled data, presenting R1 as a cost-competitive alternative to major models. Data handling notes stress that user data is not processed in China, but on European or US servers. The piece also references ongoing discussions about a potential joint venture or sale of TikTok's US business involving Perplexity, with involvement from Elon Musk, Microsoft, and other large players.
Moderate impact; modeling integration and data localization discussed, with potential TikTok collaboration/thematic market interest.
Track Perplexity Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Perplexity integrated DeepSeek's R1 for Reasoning Mode.
- Perplexity introduced Perplexity Assistant and API Sonar for embedding AI-powered search.
- R1 uses reinforcement learning and achieves high performance with minimal labeled data.
- Data is not processed in China; it is processed on European or US servers.
- There are reports of talks about a TikTok US joint venture or sale involving Perplexity, with Elon Musk and Microsoft among those connected.
Connected Companies & Entities
7 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Ecosia Drops Mistral for Chinese Open-Source AI
European search engine Ecosia has switched its AI supplier from French startup Mistral to open models, including Chinese ones like Alibaba's Qwen, Z.ai's GLM, and Moonshot AI's Kimi. CEO Christian Kroll cited disappointment with Mistral's model quality, which he says lags about a year behind competitors, and frequent server overloads. Ecosia now uses models via German platform Melious, which runs open AI models on EU servers, halving AI service costs while improving performance. The switch comes as Mistral released Large 4, its most powerful model yet, trained in Europe with open weights due in October. Despite scoring 38.4 on the Intelligence Index, it ranks eighth among open models, behind seven Chinese models. The case highlights the European AI sovereignty dilemma: top-tier open models are predominantly Chinese, even when run on European servers. Mistral itself hosts Chinese models like GLM on its neocloud platform, while its CEO Arthur Mensch defends Large 4's capabilities in areas like cyber defense.
Ecosia Switches from Mistral to Chinese AI Models
Berlin-based search engine Ecosia has dropped French AI provider Mistral and switched to open-weight models, including Chinese ones like Qwen (Alibaba), GLM (Z.ai), and Kimi (Moonshot AI), as reported by Politico. Ecosia CEO Christian Kroll was reportedly disappointed with Mistral's model quality, saying they lag behind competitors by about a year, and also questioned Mistral's sovereignty due to its reliance on international investors. Ecosia now sources models via Melious, a German platform running open AI models on European servers, cutting AI costs by half while improving performance. Mistral, meanwhile, released Large 4, a trillion-parameter model trained in European data centers, which ranks eighth among open models, with all top seven being Chinese. This highlights Europe's AI sovereignty dilemma: top open models are mostly Chinese, even as Mistral itself now hosts Chinese models like GLM on its neocloud.
OK Future's AI 'Pressure Cooker' Campaign for Goodwipes
Former MullenLowe U.S. CEO Frank Cartagena launched creative shop OK Future to test generative AI's potential for a small agency. Their first project, a spoof of OpenAI's Astra ad for personal hygiene brand Goodwipes, was produced in four days using AI tools like ArtCraft, Seedance, and OpenAI's Astra model, cutting projected production costs from $700,000. The campaign, 'Meet Asstra,' gained over 1.5 million views on Reddit. However, Cartagena described the pace as 'unsustainable' and a 'pressure cooker,' with team members working around the clock and even threatening to quit. Goodwipes' SVP of Marketing, Meredith Diehn, emphasized trust in Cartagena and the value of experimenting with AI. The article highlights the growing use of AI in creative production, with 73% of marketers using GenAI for visual content and Gartner forecasting AI software spending to reach $981 billion by 2029.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.