We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode 849: 2025 AI and Data Science Predictions, with Sadie St. Lawrence

849: 2025 AI and Data Science Predictions, with Sadie St. Lawrence

2024/12/31
logo of podcast Super Data Science: ML & AI Podcast with Jon Krohn

Super Data Science: ML & AI Podcast with Jon Krohn

AI Deep Dive AI Insights AI Chapters Transcript
People
J
Jon Krohn
S
Sadie St. Lawrence
Topics
Sadie St. Lawrence: 2024年对GPU和其他AI硬件加速器的需求将远超以往,这将为新的竞争者与英伟达等老牌巨头竞争打开大门。英伟达的股票价格大幅上涨,同时市场上也涌现了许多新的竞争者,例如AWS的Trainium和Inferentia芯片。大型语言模型作为新操作系统的预测仅部分准确,人们的习惯改变需要更长时间。大型语言模型的能力将超越简单的规模扩展,通过模仿人类的“慢思考”过程来增强逻辑推理和数学能力。未来大型语言模型的扩展将超越简单的规模化,更多地关注特定领域的模型开发,类似于生物学中动物根据环境进化出独特智力的方式。虽然大型语言模型的功能调用API整合了各种工具,但企业工具的整合程度仍有待提高,微软Copilot由于与微软套件深度集成而表现更好。复杂的代码解释器(如ChatGPT的高级数据分析功能)使业务用户能够自行执行数据分析,从而改变了传统分析师的角色,导致工作场所出现剧变。人工智能对技术领域的工作岗位影响巨大,导致许多高科技公司裁员,软件开发职位发布数量也大幅下降。人工智能对初级技术人员的影响最大,他们需要提升自身技能,特别是商业技能,才能在竞争中保持优势。开源项目,特别是Meta的Llama模型,以及最终受益的消费者,是2024年的最大赢家。2025年自主式AI将成为主导趋势,超越单一应用程序,创建能够自主处理复杂任务的专业网络,但平台间的安全性和权限仍然是一个挑战。2025年,人工智能将进一步集成到日常设备中,例如增强现实眼镜和个人电脑,但这并非所有集成都具有价值。数据科学领域的角色将不断发展和演变,需要从业者不断学习和积累新的技能,例如人工智能工程技能,这与2014年IT领域向云计算的转变类似。 Jon Krohn: 2024年,大型语言模型将成为一种新的操作系统,改变人们与机器交互的方式,减少对键盘和屏幕的依赖。大型语言模型的创新方法将致力于复制人类的“慢思考”过程,从而显著增强逻辑、推理和数学任务的能力,并可能减少对训练数据的需求。OpenAI的O1模型就是这一趋势的体现。大型语言模型的功能调用API将整合各种系统和应用程序,从而简化工作流程并形成统一的企业工具集,这在2024年通过自主式AI框架得到了体现。复杂的代码解释器(如ChatGPT的高级数据分析功能)使业务用户能够自行执行数据分析,从而改变了传统分析师的角色,导致工作场所出现剧变。人工智能对技术领域的工作岗位影响巨大,导致许多高科技公司裁员,软件开发职位发布数量也大幅下降。2024年谷歌的强势回归是年度最佳回归。谷歌凭借Gemini 2.0等产品在各个基准测试中均表现出色,重新回到了AI领域的竞争前沿。2024年OpenAI的表现令人失望,未能达到人们的预期。苹果的Apple Intelligence令人失望,未能达到预期,这可能是因为苹果公司对安全性和可靠性产品的重视。OpenAI的O1模型是2024年最令人惊艳的时刻。特斯拉的自动驾驶功能令人印象深刻,实现了儿时梦想。Waymo的无人驾驶汽车体验令人印象深刻。Waymo公司致力于打造“世界上最好的司机”,这体现了规模化机器智能的潜力。Anthropic公司及其Claude模型是2024年的最大赢家。Claude模型在调试、总结和转录等任务上表现出色,其友好的用户界面也令人印象深刻。2025年,企业对人工智能的货币化将至关重要,因为公司需要在其巨额硬件投资中获得回报。2025年,市场对人工智能工程技能的需求将超过对传统数据科学技能的需求,但这代表着角色的演变而非替代,从业者需要在现有技术基础上构建新的AI工程能力。

Deep Dive

Key Insights

What was the 'comeback of the year' in AI for 2024 according to Sadie St. Lawrence and Jon Krohn?

Google was named the 'comeback of the year' for 2024 due to its significant advancements in AI, including the release of Gemini 2.0, Willow AI Studio, and Notebook LM, which helped it regain a competitive position in the AI landscape.

Why was OpenAI considered the 'disappointment of the year' in 2024?

OpenAI was considered the 'disappointment of the year' because it failed to meet the high expectations set by its previous innovations, such as GPT-3 and GPT-4. Despite releasing the O1 model, the anticipated breakthroughs in scaling and capabilities did not materialize as expected.

What was the 'wow moment of the year' in AI for 2024?

The 'wow moment of the year' was OpenAI's O1 model, which introduced slow thinking capabilities, allowing AI to break down complex tasks into intermediate steps and validate each step, significantly enhancing logic, reasoning, and mathematical tasks.

What is the primary focus of Sadie St. Lawrence's prediction for AI in 2025?

Sadie St. Lawrence predicts that agentic AI will be the dominant trend in 2025, moving beyond single applications to create specialized networks that can autonomously handle complex tasks, though challenges around security and permissions between platforms remain.

How does Sadie St. Lawrence see AI integration into everyday devices evolving in 2025?

Sadie predicts that AI integration into everyday devices will accelerate in 2025, with advancements like augmented reality glasses offering real-time translation and more sophisticated personal computing experiences, though not all integrations will prove valuable.

What impact does Sadie St. Lawrence expect AI to have on scientific research in 2025?

Sadie expects AI-driven scientific research to expand significantly in 2025, building on current successes where AI-assisted researchers achieved 44% more new material discoveries and 39% more patents than researchers who weren't AI-assisted.

What is the key challenge in enterprise AI monetization as discussed by Sadie St. Lawrence?

The key challenge in enterprise AI monetization is ensuring profitability while addressing security and privacy concerns. Many AI projects get stuck in proof-of-concept purgatory due to difficulties in making them cost-effective and secure for production environments.

What skills does Sadie St. Lawrence predict will be in high demand in 2025?

Sadie predicts that demand for AI engineering skills will surpass traditional data science skills in 2025, though this represents an evolution of the role rather than a replacement, requiring practitioners to build on existing technical foundations with new AI engineering capabilities.

Chapters
This chapter recaps Sadie St. Lawrence's predictions for 2024, assessing their accuracy. Topics include the demand for GPUs, LLMs as a new operating system, advancements in LLM capabilities, tool consolidation via LLM APIs, and workplace upheaval.
  • Sadie's predictions for 2024 included increased demand for GPUs, LLMs transforming human-machine interaction, advancements in LLM capabilities beyond scaling, tool consolidation via LLM APIs, and workplace upheaval due to AI.
  • Three out of five predictions were deemed highly accurate.
  • The prediction about LLMs as a new operating system was considered partially accurate, with more development needed.
  • The prediction regarding workplace upheaval highlighted the impact of AI on tech jobs and the need for skill adaptation.

Shownotes Transcript

Sadie St Lawrence returns for her 4th annual prediction episode on the Super Data Science Podcast. Together with host Jon Krohn, they reflect on 2024’s most transformative trends—like agentic AI and enterprise AI monetization—and predict what's coming in 2025, from AI-driven science to the skills data scientists need to stay ahead.

Interested in sponsoring a SuperDataScience Podcast episode? Email [email protected]) for sponsorship information.

In this episode you will learn:

  • (03:30) 2024 AI trend recap

  • (19:23) Comeback of the year: Google

  • (27:29) Wow moment of the year

  • (40:20) Looking ahead to 2025

Additional materials: www.superdatascience.com/849)