Member Info for Sangijuelas1

i reckon saba got told to **** off by mv and they have 10% to sell

suits me -will buy back what i sold recently

Yeah you better sell up quick

Wonder why these have stopped especially since the recent cash from Revolut sales? And pressure from Saba

Https://www.theguardian.com/business/2026/may/29/ocado-to-take-over-all-asda-home-deliveries-from-next-year

@Rizzel

This article says orders will be through Ocado.Com like with M&S

So to order an online delivery from Asda their customers need to go to Ocado.com. Cool

Bloomberg

FT reporting that they are set to get UK banking licence, which should accelerate path to IPO. They already have application in for US licence.

GROW is not a VCT. It would be eligible for tax relief if it was and it wouldn't be able to invest in non UK companies.

There is a Molten VCT but that is a separate listing.

We want to offload the Revolut shares before IPO. If they can flog the lot at 75bn that would be a win.

Look at what happened with UIPath and Trust pilot when they listed and Molten were scrambling to sell the shares well under the IPO price

Watch it drop another 10% once they start the analyst call

We did- I used to trade it and you would often get 10% swings in a day

Taken from the comments on this article

https://archive.ph/NzJrd

Please use the sharing tools found via the share button at the top or side of articles. Copying articles to share with others is a breach of FT.com T&Cs and Copyright Policy. Email licensing@ft.com to buy additional rights. Subscribers may share up to 10 or 20 articles per month using the gift article service. More information can be found at https://help.ft.com/faq/gifting-and-sharing-an-article/what-is-a-gift-article/.

https://www.ft.com/content/92dfd571-8d34-42f1-8be8-dce126998e37

citing an invalidated precedent can lead to immediate legal malpractice, court sanctions, or a total collapse of a case's logical foundation. Unlike generic AI models that can merely summarize text, Shepard’s is built upon over 150 years of proprietary human editorial review, where thousands of legal editors manually code the "treatment" of every case (assigning signals like the Red Stop Sign for warnings or Yellow Triangle for caution). This creates an authoritative "knowledge graph" that general-purpose AI cannot replicate from public data alone, ensuring that when an agent like Lexis Protégé makes a recommendation, it is grounded in a validated historical reality that safeguards a lawyer's professional standing.

5. All of this content is proprietary and protected by strict customer agreements and legal restrictions. Access to regulated public records is governed by law and requires compliance infrastructure and exclusive relationships with courts and agencies. The content is hosted within LexisNexis’s secure, proprietary ecosystem and is not available for external training or use. Customers cannot use LexisNexis content for their own or third-party AI applications, and others cannot train AI models on this data.

this is from a broker note detailing exactly what the lexis nexis data is and how you won't access it via web search.

1.

please use the sharing tools found via the share button at the top or side of articles. copying articles to share with others is a breach of ft.com t&cs and copyright policy. email licensing@ft.com to buy additional rights. subscribers may share up to 10 or 20 articles per month using the gift article service. more information can be found at https://help.ft.com/faq/gifting-and-sharing-an-article/what-is-a-gift-article/.

https://www.ft.com/content/92dfd571-8d34-42f1-8be8-dce126998e37

lexisnexis has built the world’s most comprehensive repository of accurate and verified legal content over many decades – more than 150bn documents [2mn added daily], most of which cannot be found on the open internet. this includes primary law, secondary sources, practical guidance, expert opinion, case law, filings, court rulings, regulations, and statutes. lexis estimates it has 30% more case law than any competitor and twice as many legal briefs, pleadings, and motions. exclusive content includes matthew ****** treatises and practical guides, the largest public records collection, and the most extensive set of licensed business and news sources, with twice as many ai-licensed rights as peers. legal content is updated multiple times post-publication and includes historical documents from closed state and local offices.

2. the infrastructure behind this is industrial in scale. lexisnexis gathers primary law from over 3,000 courts, overcoming paywalls, captchas, broken feeds, and even hard-copy-only courts by employing hundreds of “runners” to physically collect documents. content arrives in thousands of formats – pdf, html, xml, tiff, word, and more – and is continuously normalized and disambiguated using specialized technology, as courts frequently update opinions and legal entities are referenced in countless ways.

3.editorial enrichment is layered on top of this raw content: headnotes, case notes, summaries, and billions of interlinkages, including shepard’s citator signals that track the legal status and treatment of cases over hundreds of years. the platform integrates authoritative brands and expert commentary from thousands of lawyers, offering practical guidance and treatises that are recognized as industry standards. lexisnexis also provides the largest collections of expert witness content, court records, unpublished case law, and analytical court material, all deeply integrated into attorney workflows.

4. shepard’s citations is the definitive legal "system of truth". it is a comprehensive citator service that allows attorneys to "shepardize" a case, which means verifying that a judicial decision or statute remains "good law" – essentially confirming it hasn't been overruled, reversed, or negatively questioned by a later court. this process is critically important because citing an invalidated precedent can lead to

Could be that. But also Anthropic completed their fundraise so they can ease up on hyping up the company.

Zebbo- this is similar to something I was just reading about how being able to be orchestration layer in tech stack is key and is literally what Anthropic are trying to do themselves

https://www.decodingdiscontinuity.com/p/285-billion-saaspocalypse-wrong-panic

Law Firm business model

Bill clients by the hour. Present itemised bill to clients including case law research cost.

They will love the sales pitch. Research that would take 20 hours will now only take 10 seconds.

Or is Anthropic going to become a global law firm employing AI lawyers.

Will still need the case law data.

This gives idea of why Relx data much of which they do own is valuable

https://www.ad-hoc-news.de/boerse/news/ueberblick/relx-plc-the-quiet-data-superpower-behind-global-decisions/68553698

I referenced Anthropic because their release of new tools is what has led to the SP collapse in the likes of Relx.

The Harvey situation is interesting. Both them and Legora are Anthropic clients but it looks like these companies will both lose business to Anthropic themselves.

A big part of Anthropic's client base are companies like Harvey that set up as an API layer and pay to access the LLM. Now Anthropic are coming with plug in tools for these companies enterprise customers.

That is an interesting situation.

Can Relx not just deploy Claude themselves to provide responses to prompts rather than licensing data to other companies to do that?