As AI continues to evolve (from machine learning to GenAI LLMs/SLMs), different industries are impacted in diverse ways. A service practice that thrives on this is management consulting, which helps enterprises make sense of these rapid changes.
Some anecdotal evidence of this management consulting traction in AI-related projects: work related to AI now accounts for a fifth of BCG’s revenue, up from zero two years ago; IBM sees more than $1 billion in sales commitments related to GenAI; Accenture booked $300 million in sales related to GenAI last year; ~40% of McKinsey’s business this year is expected to be GenAI related; and KPMG had more than $650 million in GenAI business over the past six months, up from zero a year ago.
On the proprietary, closed model front, last week, Anthropic released Claude 3.5 Sonnet (https://www.anthropic.com/news/claude-3-5-sonnet). Claude 3.5 Sonnet sets new records for foundation models on general knowledge benchmarks like MMLU (Massive Multitask Language Understanding) and coding benchmarks like HumanEval, as shown below.
On several benchmarks, it competes with or surpasses the performance of OpenAI’s GPT-4o. Anthropic has also expanded the model’s vision capabilities, enabling it to process charts and images more precisely.
Rather than which foundation model to use, the initial key decision for enterprises is whether to select an open-source or proprietary closed model.
At Alten Capital, we invest in technology services companies. Please reach out to explore potential partnerships.