IBL News | New York
OpenAI officially opened its New York office this month. It’s located at the iconic Puck Building. “This new office reflects our mission to harness the potential of GenAI to empower people to solve complex problems right here in New York,” said the company that created ChatGPT
At the same time, OpenAI announced a partnership with the New York City Economic Development Corporation (NYCEDC) through their startup accelerator program, the Founder Fellowship.
IBL News | New York
Paris-based AI startup Mistral launched a new API for content moderation last week. This addresses growing concerns about data privacy, latency, and compliance, which can prove especially attractive to European companies subject to strict data protection regulations.
This API powers the moderation service in Mistral’s Le Chat. Powered by a fine-tuned model (Ministral 8B), it can be tailored to specific applications and safety standards.
Mistral is releasing two end-points: one for raw text and one for conversational content.
The API is the same API that powers moderation in Mistral’s Le Chat chatbot platform.
It includes English, French, and German languages into one of nine categories: sexual, hate and discrimination, violence and threats, dangerous and criminal content, self-harm, health, financial, law, and personally identifiable information.
Mistral says the moderation API can be applied to raw or conversational text.
The moderation API is available through Mistral’s cloud platform, with pricing based on usage.
The release follows Mistral’s recent announcement of high-profile partnerships, including deals with Microsoft Azure, Qualcomm, and SAP.
IBL News | New York
Llama open-source AI model is increasingly used by U.S. government agencies, contractors in national security, and large companies such as Accenture, AWS, Azure, Anduril, Booz Allen, Databricks, Deloitte, IBM, Leidos, Lockheed Martin, Microsoft, Oracle, Palantir, Scale AI, and Snowflake.
Meta talked publicly about it to combat the perception that it is aiding foreign adversaries like China. Last week, Reuters reported that Chinese research scientists linked to the People’s Liberation Army (PLA), the military wing of China’s ruling party, used an older Llama model, Llama 2, to develop a tool for defense applications.
“These Llama uses will also help establish U.S. open-source standards in the global race for AI leadership,” said the company.
“Also, open-source systems have helped to accelerate defense research and high-end computing, identify security vulnerabilities, and improve communication between disparate systems.”
• Lockheed Martin incorporated Llama into its AI Factory, accelerating various use cases such as code generation, data analysis, and enhancing business processes.
Scale Ai is fine-tuning Llama to support specific national security team missions, such as planning operations and identifying adversaries’ vulnerabilities.
Meta believes “a global open-source standard for AI models is likely to emerge, as it has with technologies like Linux and Android.”
IBL News | New York
CEOs of major tech companies congratulated Donald Trump and JD Vance on their victory in the U.S. presidential election this Wednesday.
They all wished Trump success in the Oval Office and stated they looked forward to working with him.
Amazon’s Founder and owner of The Washington Post, Jeff Bezos, said in a post on X that it was “an extraordinary political comeback and decisive victory.”
Bezos recently struck a more conciliatory tone and praised Trump for his “courage under literal fire” following the attempted assassination at a Pennsylvania rally.
Apple CEO Tim Cook congratulated Trump on his victory in a post on X. “We look forward to engaging with you and your administration to help make sure the United States continues to lead with and be fueled by ingenuity, innovation, and creativity,” Cook wrote.
OpenAI CEO Sam Altman said he hopes Trump will see “huge success in the job.” In a follow-up post, he wrote, “It is critically important that the US maintains its lead in developing AI with democratic values.”
Meta CEO Mark Zuckerberg said, “We have great opportunities ahead of us as a country.”
Elon Musk, who runs Tesla, SpaceX, and Neuralink, cheered Trump’s win.
Sundar Pichai, CEO of Google parent Alphabet, congratulated Trump on his victory and said he’s committed to working with the president-elect’s administration.
Microsoft CEO Satya Nadella said, “Congratulations, President Trump. We’re looking forward to engaging with you and your administration to drive innovation forward that creates new growth and opportunity for the United States and the world.”
IBM CEO Arvind Krishna congratulated Trump on his victory in a LinkedIn post. He said the company looks forward to working with the Trump administration and Congress “to advance American leadership in AI and quantum, strengthen America’s competitiveness, and expand economic opportunity across the country.”
Cisco founder and CEO Chuck Robbins wrote in a post on X that the company looks forward to working with Trump and Congress on policies around “connectivity, innovation, cybersecurity, and more.”
Big congratulations to our 45th and now 47th President on an extraordinary political comeback and decisive victory. No nation has bigger opportunities. Wishing @realDonaldTrump all success in leading and uniting the America we all love.
— Jeff Bezos (@JeffBezos) November 6, 2024
Congratulations President Trump on your victory! We look forward to engaging with you and your administration to help make sure the United States continues to lead with and be fueled by ingenuity, innovation, and creativity.
— Tim Cook (@tim_cook) November 6, 2024
congrats to President Trump.
i wish for his huge success in the job.
— Sam Altman (@sama) November 6, 2024
It is morning in America again pic.twitter.com/GNTE0cUWoc
— Elon Musk (@elonmusk) November 6, 2024
IBL News | New York
George Washington University’s (GWU) professor Lorena A. Barba presented a white paper about a pilot project to develop and deploy AI mentors over various courses and organize a faculty community of practice to build knowledge and skills.
The pilot is supported by George Washington University’s Trustworthy AI Initiative.
It analyses how a catalog of retrieval-augmented generative AI aids for course-level learning support.
In parallel, a team of social science and education researchers will evaluate the effectiveness of the AI mentors in helping students achieve learning objectives and whether students improve their use of AI assistance over time.
GWU’s technology partners for this pilot project are ibl.ai, an edtech company that develops software products based on open-source technologies with a growing expertise in generative AI for education.
[ibl.ai is the parent company of iblnews.org]
• Generative AI for teaching and learning pilot project 2024-5—White paper PDF
IBL News | New York
The concept of AI actors, or “virtual actors,” is gaining traction as advancements in Generative AI continue to evolve.
These AI actors are entirely digital creations, capable of performing in films, commercials, video games, and other media without the need for human counterparts.
One of the latest companies in the field is aiactors.app.
In the same field, other companies are Soul Machines, Synthesia, and HeyGen.
introducing aiactors (dot) app – hire ai actors for 25X cheaper
❌ $50/video
✅ $2/videouse it instantly for product demos, announcements & more pic.twitter.com/5bNzkepjYU
— Ansh Chopra (@anshchopra_) August 19, 2024
HeyGen is absolutely great but a bit different and people don’t get the difference cause it’s hard:
So HeyGen takes a 1-3 min REAL video (or longer) of yourself speaking
It trains an AI voice of you (like ElevenLabs does)
You then write a script
It then
1) generates the AI… https://t.co/5Xgx102qoc— @levelsio (@levelsio) August 11, 2024
IBL News | New York
LLMs (large language models), still in their infancy, are becoming “more of a commodity,” said Microsoft chief executive Satya Nadella.
A common view in the industry is that it is hard to separate OpenAI’s latest GPT from Anthropic’s Claude or Google’s Gemini.
On the consumer side, Meta said last week that 500 million people are now looking at its Meta.AI at least once a month,
However, today, OpenAI’s ChatGPT’s dominance is undisputed. Its 250 million users a week and a $20 monthly subscription fee paid by a small minority result in annualized revenue of $3.6 billion.
Also, ChatGPT can point to a favored position on the iPhone, thanks to a deal with Apple.
On the downside, OpenAI, without a functional business model yet, is on track to burn through over 5 billion of cash this year,
Experts say that OpenAI’s biggest challenges are the lack of deep moats around its business and its intense competition, as the costs of querying for other LLMs have fallen rapidly.
In addition, the capabilities of open-source AI models have advanced quickly, making them viable alternatives and raising the possibility of the commodification of LLMs.
Meta’s Llama hasn’t yet become “the Linux of AI,” as Mark Zuckerberg suggested last week.
OpenAI’s latest voice-powered GPT-4o model has been credited with breaking new ground in naturalistic voice interaction, potentially opening up new consumer markets to AI.
Also, its GPT-o1 is the first model capable of breaking a complex problem.
For now, OpenAI, backed by investors at a $150 billion valuation, maintains an edge in model-building despite powerful companies in tech closing fast.
IBL News | San Antonio, Texas
The 2024 EDUCAUSE Analytics Landscape Study, presented at Educause’s annual conference last week in San Antonio, Texas, noting that higher education institutions understand the far-reaching impact of data and analytics on their organizations.
In fact, most institutional leaders (79%) are interested in or fully committed to analytics.
Regarding AI, Educause found that only a small number of institutions are using AI tools to support analytics, and these tools are slowly gaining traction.
Other findings include:
• Analytics are most frequently used in admissions/enrollments and compliance with accreditation standards and regulatory requirements.
• Smaller institutions have lower access to resources and support for their use of analytics.
• Institutions are lacking in dedicated analytics leadership and staff positions. Most said their institution does not have a chief analytics officer (69%) or a chief data officer (57%).
• Analytics is a strategic priority, but governance is lacking. Most respondents said that analytics is considered a strategic priority (69%) and necessary for effective decision-making at their institution (67%).
“Success hinges on data: data quality, data integration, data governance, data management, and data literacy,” said Susan Grajek, Vice President of Partnerships, Communities, and Research at Educause.
IBL News | New York
In a direct challenge to Google, OpenAI significantly improved its web search functionality on its ChatGPT, adding links to relevant sources in the queried results. A simple web search icon activates the process.
These search results include news, blog posts, stock quotes, sports scores, and other timely information.
OpenAI has partnered with news and data providers to provide this up-to-date information. Those media partners include Associated Press, Axel Springer, Condé Nast, Dotdash Meredith, Financial Times, GEDI, Hearst, Le Monde, News Corp, Prisa (El País), Reuters, The Atlantic, Time, and Vox Media.
The company made this announcement yesterday.
Search is available only to Plus and Team paid users and waitlist users. In “the next few weeks,” Enterprise and Edu users will have access. OpenAI will roll out this feature to all Free users over the coming months.
The functionality is available at chatgpt.com and ChatGPT’s desktop and mobile apps.
The search model is a fine-tuned version of GPT-4o, post-trained using novel synthetic data generation techniques, including distilling outputs from OpenAI o1-preview.
“Thanks to feedback from the SearchGPT prototype, we brought the best of the SearchGPT experience into ChatGPT. We plan to keep improving search, particularly in areas like shopping and travel, and leverage the reasoning capabilities of the OpenAI o1 series to do deeper research. We also plan to bring our new search experience to Advanced Voice and canvas, as well as to Free and logged out users in the future,” announced OpenAI.
IBL News | New York
Anthropic AI introduced a new built-in feature for Claude.ai this week that enables this chatbot to write and run JavaScript code.
This analysis tool can analyze and render interactive visualizations from CSV files and PDFs, conduct analysis, and do complex math, producing real-time insights. It iterates on different ideas before sharing an answer.
“The ability to process information and run code means you get more accurate answers—building on Claude 3.5 Sonnet’s state-of-the-art coding and data skills,” said the company.
Anthropic has offered this feature before, but since it lacks a mechanism to verify the results mathematically, the answers aren’t always accurate.
Anthropic provided a few examples of where this tool might be useful:
Similarly, OpenAI’s models can write and execute code via a feature called Advanced Data Analysis.
Google offers a comparable feature called Code Execution on its Gemini models. It runs and generates Python code that learns iteratively from the results.
[Disclosure: ibl.ai, the parent company of iblnews.org, a partnership with Google Cloud]
As a marketing promotion, OpenAI will offer all 2025 $5,000 on API credits to help them innovate and grow. OpenAI is contributing $375,000.
The Founder Fellowship has supported 250 entrepreneurs across 168 companies, with founders raising over $22 million in venture capital and grant funding.