Being able to clearly describe a problem and work with the AI to design a solution, prioritise what to put the AI to work on, set up good harnesses so the quality of the output is kept high, figure out what parallelises well and what’s going to set off agents that are stepping on each others toes… all of this needs experience and judgement and delegation and project organisation skills.
AI is supercharging tech leads. Beginners might be able to skill up faster, but they’re not getting the same results.
Google Search has no stickiness and they managed to build a behemoth.
ChatGPT is a great product, but the lack of stickiness comes into play because there are many viable alternatives.
They’re all going to have to monetise the consumer segment at some stage, and I think that’s likely to be via ads on a freemium tier in most instances.
The number of ReDoS vulnerabilities we see in Dependabot alerts for NPM packages we’re only using in client code is absurd. I’d love a fix for this that was aware of whether the package is running on our backend or not. Client side ReDoS is not relevant to us at all.
I dunno, that seems a bit narrow minded to me. You're making an assumption about talking to AI being a worse experience than talking to a person (which is frequently _terrible_).
What if you were able to get helpful support, 24/7/365, with no time waiting in a queue, in your own language (regardless of the service provider's location and 'native' language support)? And the company was able to provide the product and support for it cheaper, resulting in less cost to you?
My take is that they’re not necessarily harmful in and of themselves, but it’s absolutely harmful to think that this is the way to grow and get traction. It’s not a repeatable approach and it’s likely to pump some top of funnel metrics temporarily without having meaningful impact to the bottom line.
This can be very distracting if you’re pursuing it intentionally and treating ‘going viral’ as a prerequisite to success.
Today they have announced further, related vulnerabilities, and if you're running your own instance you should patch again, or disable your instance until you have a chance to do so.
The vulnerabilities allow an unauthenticated attacker to run arbitrary commands with the same privileges as the Metabase server on the server you are running Metabase on. This would allow arbitrary querying of any database that Metabase is connected to.
UsabilityHub | Engineering | Remote (Australia or New Zealand) | Full-time, 9 day fortnight, or 4 day week | https://usabilityhub.com/careers
Hey HN, Nick here, CTO and co-founder at UsabilityHub.
We're looking to hire a couple of engineers, with some flexibility on seniority level. We're a remote first and cross functional team, and UsabilityHub is a bootstrapped and successful business that's been focused on sustainable growth over the last ten years.
The team is small, close knit, and has some really excellent engineers and designers. The focus of these roles is building new customer facing functionality - we're effectively building three new products this year so there's lots of greenfields work to do.
The big pieces of our tech stack are Ruby, Rails, Typescript, Postgres. We're not fussy if you haven't used some or any of those, but prefer T-shaped developers that are keen and able to pick new things up quickly and contribute throughout the stack.
For the three roles we have open, starting compensation ranges from $120k AUD to $160k AUD (or NZD equivalents) plus ESOP and profit share. We review this every six months and adjust upwards based on performance in the role.
There's lots more info about these specific roles on our careers page at https://usabilityhub.com/careers. Feel free to reach out to me directly with any questions (email is in profile).
Some kind of concurrency bug in a library they were using to retrieve cached data from Redis led to this leak.
> We took ChatGPT offline earlier this week due to a bug in an open-source library which allowed some users to see titles from another active user’s chat history. It’s also possible that the first message of a newly-created conversation was visible in someone else’s chat history if both users were active around the same time.
Surely they can aggregate this across all customers though.
If Twilio cops an unexpectedly high settlement for sending an SMS to +1234567890 in January, can they assume that a separate customer sending an SMS to that number in February will end up in the same boat?
I'd be very surprised if the toll fraudsters weren't using the same numbers to hit multiple Twilio accounts.
Yes, but other queries (any aggregate queries that don't join the soft deleted table, any joins to other tables) will now return rows that would have been deleted under hard deletion with cascade.
> Performance only matters if you are in competition with others.
I'm not sure that's right. There's a strong link between performance and repeat usage. Even in the complete absence of competition, Google search being fast could result in more searches per user being conducted and more revenue as a result.
Being able to clearly describe a problem and work with the AI to design a solution, prioritise what to put the AI to work on, set up good harnesses so the quality of the output is kept high, figure out what parallelises well and what’s going to set off agents that are stepping on each others toes… all of this needs experience and judgement and delegation and project organisation skills.
AI is supercharging tech leads. Beginners might be able to skill up faster, but they’re not getting the same results.