Guillaume Loret
@yom
Followers
426
Following
880
Media
124
Statuses
7K
🇪🇺 in the 🇺🇸 (background picture by Nicolas Raymond: https://t.co/ULMIqvRPI8)
Washington, DC, USA
Joined June 2007
LangChain is moving away from Retrieval Augmented Generation (RAG) and towards agentic search. [ https://t.co/x0jvCNiERC
0
0
0
Ce visuel de Bloomberg est excellent (et un peu terrifiant aussi)... C'est l'explication de ce qu'est une bulle et là il s'agit d'une bulle dans l'IA. Quand les investissements deviennent circulaires, sans irriguer l'économie réelle - OpenAI achète des puces Nvidia - Nvidia
40
179
529
His journey in AI: finding clients beats building. AI accelerates, niches vanish, judgment & personal brand are the real moats. Be realistic. https://t.co/MdPaFlhAEv
reddit.com
Explore this post and more from the AI_Agents community
0
0
0
Someone at a16z claimed a few weeks ago that 80% of Bay Area startups are building on Chinese open source models. The graphic below shows Chinese model downloads exceeding US models on HuggingFace.
67
352
2K
60% of CS PhDs in the US are not American. 53% of CS Masters grads in the US are not American. BigTech cos have <15% employees on H-1B. BigTech hiring isn't necessarily biased against Americans; most Americans just don't study advanced CS!
310
309
2K
Constrained Transformation: Cities can develop new industries and abandon old ones, but they must do so while maintaining coherence. Even cities like Pittsburgh and Boston that shifted from heavy manufacturing to high-tech had to find paths that preserved their economic coherence
1
0
0
Size Matters: Larger cities are consistently less coherent, with coherence declining about 4% for each doubling of population. Bigger cities can sustain more diverse, loosely connected activities.
1
0
0
A highly coherent city like Detroit (during its automotive heyday) focuses on closely related industries, while a less coherent city like New York spans many unrelated sectors
1
0
0
Three ingredients matter: 1. Variety – how many different activities exist in the city. 2. Balance – how evenly workers or firms are spread across those activities. 3. Relatedness – how similar or connected those activities are to one another.
1
0
0
For over 170 years, as cities evolve from craftsmanship to technology, they maintain a consistent "economic coherence." It measures how well the mix of jobs, industries, and technologies in a city fit together.
1
0
0
Many English phrases treat thinking as walking:
upworthy.com
You'll never see "by the way" and "of course" the same way again.
0
0
0
Recipients must then dissect, verify, add missing context, or re-create the work. Avoid blanket “use AI” directives. Encourage employees to experiment with AI to see where it fits or not. Use it as a tool to enhance creativity. Insist on achieving goals more than just doing AI.
1
0
0
Research suggests that most firms implementing generative AI experience more "workslop" than real value. "Workslop" refers to AI-generated content that appears polished but lacks critical context or depth, leading to increased workload for colleagues.
1
0
0
Reverse-engineer LLMs to extract structured datasets from Qwen3, GPT-OSS & Llama3 using hierarchical topic decompression https://t.co/SQcvY2BBTr
scalarlm.com
Large Language Models compress massive amounts of training data into their parameters. This compression is lossy but highly effective—billions of parameters can encode the essential patterns from...
0
0
0
US & EU far more distrustful of self-driving cars than global average. US most skeptical; global trust higher (driven by India, Pakistan, China). https://t.co/6Q2Eo9KaPq
visualcapitalist.com
Global survey data shows emerging markets have the most trust in self-driving cars, while the U.S. and Europe remain skeptical.
0
0
0
Build long-term, sustainable teams - business is a marathon, not a sprint. https://t.co/t9GgvcNuhV
fortune.com
The entrepreneurs who create lasting value—and the exits that matter—aren't the ones burning themselves out in hundred-hour weeks.
0
0
0