Skip to content

A practical perspective for individuals

Less monitoring. More value.

We turn new tools and features into one question: what can you do with them today, without unnecessary hype?

  • AvailabilityIs it actually out?
  • BenefitWhat work could this improve?
  • First stepWhat is worth verifying?
939 practical events in the selection 72 active sources Updated

Practical radar

What deserves your attention

Browse the archive →
worth noting Other

VŠEM abolished traditional bachelor's theses, students now defend practical projects

In 2024, Vysoká škola ekonomie a managementu (VŠEM) became the first school in the Czech Republic to abolish mandatory bachelor's theses and replace them with the defence of practical projects – a response to the fact that language models can both write and evaluate texts.

What you get out of it

For students and junior employees (e.g. lawyers, programmers), the VŠEM case signals that AI can now handle writing and research tasks that previously demonstrated qualifications at a comparable quality and much faster, so practical application and the ability to adapt are valued more.

1 source
Details and sources →
worth noting AI agents

VS Code is phasing out the Local harness for GitHub Copilot; the Copilot harness is the replacement

VS Code 1.146 has marked the Local harness for the GitHub Copilot agent as deprecated; it will be removed in February 2027 and replaced by the built-in Copilot harness, which has been the default since version 1.143.

What you get out of it

Developers using the Local harness in VS Code must switch to the Copilot harness by February 2027 and verify in advance that their custom settings and workflows work, since they will not be migrated automatically.

✓ official
Details and sources →
worth noting Security

False copyright strike by Media Nov Records blocked musician Samantha Shokin's video on YouTube

YouTube took down musician Samantha Shokin's video featuring a cover of a publicly available Bulgarian folk song following a false copyright claim by the company Media Nov Records. YouTube rejected both of her counterclaims, the company does not respond to contact attempts, and its website appears to be AI-generated.

What you get out of it

Musician Samantha Shokin shows that YouTube's automated copyright claims system can block even a publicly available work with no actual rights holder, and that an appeal (counterclaim) can be rejected with no way to contact a human on either YouTube's side or the side of the company that filed the claim.

1 source
Details and sources →
worth noting Business and investment

FTMO plans to open 300 new jobs in the Czech Republic over 14 months

FTMO has announced plans to open 300 new positions in the Czech Republic over 14 months, 200 of them technical, including AI specialists and product engineers. According to the company, AI saves work, but decision-making and responsibility remain with people.

What you get out of it

Candidates for technical positions (developers, AI specialists, product engineers) have access in Prague to a concrete offer of hundreds of positions and an accelerated five-hour recruitment process instead of the traditional three rounds of interviews.

1 source
Details and sources →
worth noting Security

Research describes a vulnerability of vision-language models in image downscaling involving a hidden instruction

Research on arXiv describes a vulnerability called Scale-Induced Semantic Shift: a hidden instruction in an image is not triggered in vision-language models at 512×512, but activates after downscaling below 128×128. The ScaleGuard defense reduced the attack success rate from 61.4% to 24.3%.

What you get out of it

Developers and users of tools built on vision-language models should be aware that a text instruction hidden in an image may not manifest under standard display, but can be activated after aggressive downscaling (below 128×128 pixels) during automatic processing.

1 source
Details and sources →
worth noting AI agents

United Airlines bans AI agents from booking flight tickets

United Airlines will not allow personal AI agents to buy tickets due to concerns about leaks of customer data and possible circumvention of federal rules. Agents may only search for and display flights; the purchase must be completed by the user themselves.

What you get out of it

Users who want to let an AI agent handle the entire ticket purchase with United will run into a limitation – the agent may only search for and display the flight, while the user must complete the order themselves.

1 source
Details and sources →
worth noting Productivity

AWS releases UAEF, a framework-agnostic evaluation system for AI agents

AWS has released UAEF, a framework-independent tool for evaluating AI agents with 36 metrics across 7 dimensions. It normalizes outputs from different frameworks (LangGraph, LangChain, Strands Agents, Bedrock AgentCore) into a single format and was tested by 12 organizations.

What you get out of it

Developers of AI agents get a unified way to measure quality (36 metrics across 7 dimensions) without having to build their own evaluation logic separately for each framework, making it easier to assess whether a change to a prompt, model, or agent tools actually improved things.

1 source
Details and sources →
worth noting Security

A series of cyberattacks on open-source AI platforms increases pressure on their security

Attacks on OpenCode, Hugging Face, and Fireworks stole API keys, model weights, and billing data. According to its chief policy officer, Hugging Face is strengthening monitoring and token rotation so that the open AI community doesn't have to close itself off.

What you get out of it

Developers and hobbyists contributing to open-source AI projects may face stricter identity verification or restricted access to features if platforms continue to defend against attacks; users of third-party API keys risk their misuse in similar incidents.

1 source
Details and sources →
worth noting Security

Test shows AI-generated scientific papers can bypass IEEE's review process

Journalist and mathematician Benjamin Skuse tried to push AI-generated scientific papers through IEEE's review process. Human review and integrity-checking tools, according to the source, are so far keeping up with freely available AI tools, but more advanced LLMs are harder to detect.

What you get out of it

Readers and professionals who rely on peer-reviewed publications or Wikipedia when making decisions should bear in mind that even such vetted content may be AI-generated without human editing, and where possible should verify key facts against other sources.

1 source
Details and sources →
OpenAI worth noting Regulation and law

OpenAI will respect the robots.txt signal for AI training preferences

OpenAI will respect a shared signal in robots.txt that allows website operators to express preferences regarding the use of content for AI training. The signal is meant to apply across multiple AI providers.

What you get out of it

People running their own website or blog will have a simpler tool available to indicate that they do not want their content used to train AI.

1 source
Details and sources →
worth noting Coding

AWS describes expanding QA Studio with application management, test metrics and team permissions

The QA Studio reference solution assigns tests to applications and teams, records results in Amazon CloudWatch and calculates a test health score. The expansion described also includes access roles and application sharing between teams.

What you get out of it

A tester using QA Studio can go directly from the overview of recent failures to the execution log and session recording, which provide evidence for investigating the error.

1 source
Details and sources →
OpenAI worth noting Productivity

OpenAI investigates degraded performance in ChatGPT Work

OpenAI reports degraded performance in ChatGPT Work and is investigating potential operational issues. It is still assessing the impact on users; the incident remains under investigation.

What you get out of it

When using ChatGPT Work, allow for the reported performance degradation; the company is still assessing the specific impact on users.

✓ official
Details and sources →

When personal experience proves useful

Does it make sense for a team too?

Look at the same topics through the lens of processes, risks, and implementation within a company.

Go to the company radar →